Research ArticleImprovement of High-Power Three-LevelExplosion-Proof Inverters Using Soft Switching ControlBased on Optimized Power-Loss Algorithm

Shi-Zhou Xu and Feng-You He

Department of Information and Electrical Engineering China University of Mining and TechnologyNo 1 Daxue Road Xuzhou Jiangsu 221116 China

Correspondence should be addressed to Feng-You He hfy cumt263net

Received 25 December 2014 Revised 3 February 2015 Accepted 4 February 2015

Academic Editor Ahmed El Wakil

Copyright copy 2015 S-Z Xu and F-Y HeThis is an open access article distributed under theCreativeCommonsAttribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

The high-power three-level explosion-proof inverters demand high thermal stability of power devices and a set of theories andmethods is needed to achieve an accurate power-loss calculation of power devices to establish heat dissipationmodel andultimatelyto reduce the power loss to improve thermal stability of system In this paper the principle of neutral point clamped three-level(NPC3L) inverter is elaborated firstly and a fourth-order RC equivalent circuit of IGBT is derived on which basis the power-lossmodel of IGBT and the optimized maternal power-loss thermal model using an optimized power-loss algorithm are establishedSecondly in accordance with the optimized maternal power-loss thermal model the generic formulas of power-loss calculationare deduced to calculate the power-loss modification values of NPC3L and soft switching three-level (S3L) inverters which willbe the thermal sources during thermal analysis for maternal power-loss thermal models Finally the experiment conducted on the21MW experimental platform shows that S3L inverter has the same excellent output characteristics with NPC3L inverter reducesthe power loss significantly by 213W in each half-bridge and decreases the temperature by 10∘C coinciding with the theoreticalcalculation which verifies the accuracy of optimized power-loss algorithm and the effectiveness of the improvement

1 Introduction

In explosion-proof inverters field the NPC3L inverter isone of the most mature facilities of high-power three-levelinverters at present [1] The high-power explosion-proofinverters have the features of high current flowing throughthe main circuit power devices great power losses andhigh reliability requirement What is more from the view ofapplications there is a serious problem that the power lossof inverter power devices is too great which will cause ahigh failure rate of inverter power devices and poor thermalstability of the whole system In order to improve the existingNPC3L inverters there are three issues to be addressed Thebasal one is the accurate power-loss calculation of powerdevices and it is the premise of thermal analysis and con-verter improvement The second one is a general power-losscalculation and analysis theory of three-level inverter actingas evaluation criteria to predict the results of improvements

Finally a new topology should be introduced to reduce thepower loss effectively

Generally speaking accurate power-loss calculation canfigure out the existing power-loss values of three-level invert-ers which will be a thermal source during the thermalanalysis of inverter system The inverter temperature rise ismainly caused by conduction loss and switching loss of powerdevices while the conduction and switching characteristicsof the power devices are very sensitive to temperature socalculating the power loss of the device accurately is thefoundation to optimize the design of inverters Currentlythere are many researches on power-loss calculation andthermal analysis for single IGBT module and two-levelinverters [2ndash6] However three-level and two-level invertercurrents are essentially different in the flow paths and theirlosses of power devices are of huge difference The factthat the literature [7ndash10] did not consider the impact ofjunction temperature of power devices on power losses is

Hindawi Publishing CorporationJournal of Electrical and Computer EngineeringVolume 2015 Article ID 571209 14 pageshttpdxdoiorg1011552015571209

2 Journal of Electrical and Computer Engineering

the main reason causing errors between their theoreticalcalculations and experimental results where Dieckerhoff etal [10] considered that the switching power loss of powerdevice has a linear relationship with its withstanding voltagewhile this assumption is approximately valid only in plusmn20range of the test voltage A much accurate losses calculationand heat dissipation method was introduced in [11] but itdid not take all the thermal sources in consideration whichhas an effect on the power devices and thermal analysis Inthe literature [12] the transient modeling of loss and thermaldynamics in power semiconductor devices is analyzed whileit needs to improve the model by considering the peripheralcircuits Several soft switching inverter types and controlmethods are proposed in [13ndash18] where the S3L inverter in[18] has a much more significant effect on the reduction ofpower losses It is the accurate thermal analysis methods ofinverter system that can analyze the inverter temperaturequantitatively providing references for inverter improve-ments [19 20] In the existing loss calculation studies of three-level inverters it lacks a system of theories and methodsto provide theoretical support for the improvements Beforea new three-level topology improving the NPC3L inverterit is necessary to apply a common theoretical calculationand method to anticipate its advantages The S3L inverterproposed in the literature [18] holds the viewpoint that it canreduce the power loss in terms of the NPC3L inverter underthe same conditions but there is no quantitative experimentaltemperature to support it and demonstrate its effectiveness ofimprovement

For the above reasons a general power-loss calculationmethod of three-level inverters was established in this paperbased on the optimized power-loss algorithm in Section 2with which an accurate power-loss calculation and per-formance evaluation approach of three-level inverters wasproposed To improve NPC3L inverter the S3L inverterworking principlewas elaborated in Section 3 andput into thegeneral approach mentioned above What is more accordingto this approach it is expected in Section 4 that the S3Linverter has the same excellent output characteristics withNPC3L inverter and it can reduce power loss by 213Wbringing in a 10∘C decrease in temperature intuitively In thesame section the experiment results support the validity ofthe theoretical prediction Finally Section 5 concludes thispaper

2 General Optimized Power-Loss AlgorithmBased on NPC3L Inverter

The main circuit topology of NPC3L inverter is shown inFigure 1

Each leg has four IGBTs labeled 1198791198941 1198791198942 1198791198943 and 119879

1198944

(where 119894 represents one phase of 119886 119887 and 119888 phases and eachIGBT has one antiparallel diode labeled 119863

1198941 1198631198942 1198631198943 and

1198631198944 resp) and two clamping diodes labeled 119863

1198945and 119863

1198946

[11 21]At present the IGBT device is a power module packaged

by one IGBT and a fast recovery antiparallel diodeTherefore

Udc

P

C1

O

N

C2

Ta1

Ta2

Ta3

Ta4

Da1

Da2

Da3

Da4

Da5

Da6

ab c

UVW

Tb1

Tb2

Tb3

Tb4

Db1

Db2

Db3

Db4

Db5

Db6

Tc1

Tc2

Tc3

Tc4

Dc1

Dc2

Dc3

Dc4

Dc5

Dc6

Figure 1 The main circuit topology of NPC3L inverter

its total power loss is composed of these two parts expressedas follows

119875mod = 119875119879+ 119875119863 (1)

The equivalent structure model of power device and heatsink is shown in Figure 2

It can be seen fromFigure 2 that the wholemodel consistsof four conductive layers and therefore if we consider thepower device and the heat sink as a maternal model the fourconductive layers would be four submodels On this basis thethermal resistance and heat capacity of the four submodelscan be calculated at first respectively and then all the foursubmodelsrsquo thermal resistance and heat capacity constitutethe total thermal resistance and heat capacity of the wholemodel

The calculation formula of thermal resistance 119877th isdescribed as follows

119877th =Δ119879

119875

=Δ119879

119876Δ119905

(2)

where Δ119879 is the temperature increase of submodel 119875 and Δ119905are the heat flow and time period respectivelyThen the totalthermal resistance of maternal model is shown as

119877th-total = 119877th1 + 119877th2 + 119877th3 + 119877th4 (3)

The calculation formula of thermal capacity 119862th is deliv-ered as follows [20 21]

119862th =119876

Δ119879

(4)

The total heat capacity of the maternal model can bewritten as [20 21]

119862th-total = 119862th1 + 119862th2 + 119862th3 + 119862th4 (5)

Thus thematernal model can be replaced alternatively bya fourth-order RC circuit shown in Figure 3

Journal of Electrical and Computer Engineering 3

Chip

Solder

Solder

Copper layer (Al2O3AlN)Ceramics

Copper layer

Board

Heat sink

Submodel 1

Submodel 2

Submodel 3

Submodel 4

Junctiontemperature (Tvj)

Chip-shelltemperature (ΔTjc)

Shell temperature (Tc)Shell heat sink

temperature (ΔTch)

Chip-case thermalresistance Rthjc

Pin-IGBT Pout-IGBT

Heat sinktemperature (Th)

Heat sinkambiance (ΔTha)

Ambienttemperature (Ta)

Case heat sinkRthch

Heat sinkambiance Rthha

Figure 2 Equivalent structure model of power device and heat sink

TvjT Rth1 Rth2 Rth3 Rth4

Cth1 Cth2 Cth3 Cth4PT

Tc

(a)

Cth1 Cth2 Cth3 Cth4PT

TvjTRth1 Rth2 Rth3 Rth4 Tc

(b)

Figure 3 The fourth-order RC thermal resistance equivalent circuits of IGBT (a) Partial network (b) continuous network

TvjT

TvjD

RthjcT

RthjcDRthca

Rthch

RthhaPT PD

Tc

Th

Ta

Figure 4The thermal resistance equivalent circuit of IGBTmodulein steady state

According to the partial network structure of Figure 3(a)the IGBT thermal resistance can be derived as [20 21]

119885thjc119879 =119899

sum

119894=1

119877th119894 (1 minus eminus119905120591119894) (119894 = 1 2 3 4) (6)

where 120591119894is the RC time constant of each layer

The thermal equivalent circuit of IGBT module in steadystate is shown in Figure 4

As a switching device the IGBTrsquos power loss 119875119879

isprimarily composed of conduction loss 119875con119879 and switchingloss 119875sw119879 namely [3ndash5]

119875119879= 119875con119879 + 119875sw119879 (7)

The conduction resistance initial saturation voltage andconduction loss of IGBT can be expressed respectively asfollows [11]

119903119879= 119903119879-25∘C + 119870119903119879 (119879vj119879 minus 25

∘C)

V0119879

= V0119879-25∘C + 119870V0119879 (119879vj119879 minus 25

∘C)

119875con119879 = V0119879119868 + 1199031198791198682

(8)

where 119903119879-25∘C and V

0119879-25∘C are the conduction resistanceand initial saturation voltage of IGBT with the junctiontemperature at 25∘C 119870

119903119879and 119870V0119879 are the initial saturation

voltage and conduction resistance temperature correctionfactor of IGBT 119879vj119879 is the junction temperature of IGBT 119868is the instantaneous current flowing through the IGBT

Combine the three-level working principle and optimizedIGBT power-loss model and the average conduction and

4 Journal of Electrical and Computer Engineering

switching losses formula ofT1in amodulation voltage period

will be as follows [11 22]

119875npcspwmcon1198791

= 1198910

119902

sum

119896=119901

(V01198791

+ 1199031198791119868119871(119896)) 119868119871(119896) 120591 (119896) 119879s

119875npcspwmsw1198791

= 1198910

119902

sum

119896=119901

119864sw1198791 (119868119871 (119896))

(9)

where 1198910is the frequency of modulation voltage 120591(119896) is

the 119896th duty cycle of switching period 119868119871(119896) is the average

load current of the 119896th switching period 119901 and 119902 representthe sampling periodrsquos beginning and end of 119879

1during one

modulation period respectivelyGenerally speaking when the carrier ratio is large

enough the discrete power-loss formula can be transformedinto a continuous integral form and the average conductionlosses and switching losses of 119879

1can be expressed as [22]

119875npcspwmcon1198791

=1

2120587

int

120587

120593

(V01198791

+ 1199031198791119894119871(120572)) 119894119871(120572)119863 (120572) 119889120572

=

119898V01198791

119868119898

4120587

((120587 minus 120593) cos120593 + sin120593)

+

11989811990311987911198682

119898

4120587

(1 +4

3

cos120593 + 1

3

cos 2120593)

119875npcspwmsw1198791

=1

2120587

int

120587

120593

119891sw119864sw1198791 (119894119871 (120572)) 119889120572

=119891sw2120587

(119860 sw1198791198682

119898

1

2

(120587 minus 120593 +1

2

sin 2120593)

+ 119861sw119879119868119898 (1 + cos120593) + 119862sw119879 (120587 minus 120593) )

sdot (119880dc2

119880base)

119863sw119879

(

119879vj1198791119879base

)

119870sw119879

(10)

In accordance with the same calculation principle asT1rsquos the conduction losses and switching losses (or reverse

recovery losses) of119863111987921198632 and119863

5in the same half-bridge

leg will be 119875npcspwmcon1198631 119875npcspwm

rec1198631 119875npcspwmcon1198792

119875npcspwmsw1198792

119875npcspwmcon1198632

119875npcspwmrec1198632 119875npcspwm

con1198635 and 119875npcspwmrec1198635

Some explanatory notes in expressions (8)sim(10) are asfollows

V0119879119909

= V0119879-25∘C + 119870V0119879(119879vj119879119909 minus 25

∘C) denotes the 119909thIBGTrsquos initial saturation voltage119903119879119909

= 119903119879-25∘C + 119870

119903119879(119879vj119879119909 minus 25

∘C) represents the 119909thIGBTrsquos conduction resistance119879vj119879119909 means the 119909th IGBTrsquoS junction temperature

V0119863119909

= V0119863-25∘C + 119870V0119863(119879vj119863119909 minus 25

∘C) indicates the119909th fast recovery diodersquos initial saturation voltage119903119863119909

= 119903119863-25∘C + 119870119903119863(119879vj119863119909 minus 25

∘C) stands for the 119909thfast recovery diodersquos conduction resistance

TvjT Rth1

Cth1

Rth2

Cth2

Rth3

Cth3

Rth4

Cth4PT

Tc

CaddRadd

Figure 5 The equivalent circuit of optimized power-loss model

119879vj119863119909 means the 119909th fast recovery diodersquos junctiontemperature what is more 119909 = 1 2 5 [22]

In accordance with the above step the power losses ofheat sink and IGBT maternal model can be calculated How-ever there may be buffer circuit or something like that in theperiphery around the power devices of different three-leveltopologiesThatmeans part of the power difference (119875in-IGBTminus119875out-IGBT) flowing through the IGBT model is dissipated inthe IGBT model and some other part is consumed by theperipheral circuits namely

119875in-IGBT minus 119875out-IGBT = 119875IGBT + 119875add (11)

Hence it is necessary to optimize the maternal modeland the equivalent circuit of optimized maternal model isshown in Figure 5

According to the equivalent circuit of optimizedmaternalmodel the total power-loss equation of half-bridge leg can bemodified as follows

119875total = 1198751198791+ 1198751198631+ 1198751198792+ 1198751198632+ 119875add (12)

The general power-loss calculation of half-bridge leg ofthree-level inverters using SPWMmodulation algorithm in amodulation period will be the one as follows [22 23]

119875total = 119875npcspwmcon1198791

+ 119875npcspwmsw1198791

+ 119875npcspwmcon1198631

+ 119875npcspwmrec1198631 + 119875

npcspwmcon1198792

+ 119875npcspwmsw1198792

+ 119875npcspwmcon1198632 + 119875

npcspwmrec1198632 + 119875add

(13)

where the SPWM modulation algorithm can be replacedby the actual algorithm but the power-loss calculation hasthe same process based on optimized power-loss algorithmand maternal model and all we should do is to changethe variables consistent with the algorithm we are goingto use In addition during the analysis of maternal mod-ule in Section 4 it is necessary to modify 119875add by con-sidering precharge current-limiting resistor balanced resis-tors absorption capacitance DC-link capacitors and bufferdevices of S3L inverter

3 S3L Inverter Principle

As shown in Figure 6 one full-bridge leg topology of S3Linverter contains four IGBTs (1198791015840

1198861

sim 1198791015840

1198864

) four diodes (11986310158401198861

sim

1198631015840

1198864

) snubber inductor snubber capacitances11986210158401

and11986210158402

andfour snubber diodes 119863

1198861198791sim 1198631198861198794

where the latter fourconstitute the snubber circuit [14]

Journal of Electrical and Computer Engineering 5

Table 1 Switching states of S3L inverter

Switching state + 0 minus

1198801015840

load +119880119889

2 0 minus119880119889

2

Conduction 1198791015840

1198861

or11986310158401198861

1198791015840

1198862

11986310158401198862

or 11987910158401198863

11986310158401198863

1198791015840

1198864

or11986310158401198864

1198791015840

1198861

ON OFF OFF1198791015840

1198862

OFF ON ON1198791015840

1198863

ON ON OFF1198791015840

1198864

OFF OFF ON

Ud2

Ud2

+

minus

P998400

0998400

N998400

DaT1

DaT2

DaT3

DaT4

L

Load

C9984001

C9984002

T998400a1

T998400a2

T998400a3

T998400a4

D998400a1

D998400a2

D998400a3

D998400a4

U998400load

Figure 6 One full-bridge leg topology of S3L inverters

S3L three-level inverter switching state and commutationprocess are shown in Tables 1 and 2

For zero load current commutation process it can beconsidered as three special cases specified in Table 3

Each of these commutation processes is slightly differentand therefore only the 119879

1015840

1198861

rarr 1198631015840

1198863

1198791015840

1198863

was chosen todescribe the working details as an example In order tofacilitate the analysis the load current in the commutationprocess is supposed to be constant substantially and its pathis marked in red

Before the commutation process begins 11987910158401198861

carries thepositive load current 119868Load and 119879

1015840

1198863

is switched on (butdoes not carry current because of diode 1198631015840

1198863

) 11987910158401198862

and 1198791015840

1198864

are switched off The output terminal is connected to thepositive terminal of the input DC voltage The capacitor 1198621015840

1

is discharged the capacitor11986210158402

is charged to minus119880119889 The current

in the snubber inductor is zero (Figure 7(a)) [14]The 1198791015840

1198861

rarr 1198631015840

1198863

1198791015840

1198863

commutation process starts as soonas 11987910158401198861

is switched off when 1198791015840

1198862

is switched off and what ismore 1198791015840

1198863

and 11987910158401198864

remain switched on and off respectively Inaccordance with the different current path and IGBT actionsequences the whole process can be divided into two periods

(1) 1199050le 119905 lt 119905

1Period Two current loops are generated

during this stage One of them is the oscillating current loopconstituted by 1198621015840

2

11987910158401198862

11986310158401198862

119871 1198801198892 and 119863

1198861198794 the other is

the load current loop generated by the load current flowingthrough 1198621015840

2

load midpoint 0 1198801198892 and 119863

1198861198794 As shown in

Figure 7(b) the two current paths overlap each other It isnoteworthy that the current flow decreases rapidly to zeroand the rising slope of the voltage both ends is limited to asmall amplitude so that the power loss is correspondingly

small At this time the switching-off process of 11987910158401198861

is the so-called soft switching

(2) 1199051le 119905 lt 119905

2Period The first period of commutation

process comes to an end when 1198621015840

2

discharges and 1198631015840

1198864

starts conducting At the same time the current flowingthrough inductor 119871 reduces to 0 and what is more 119879

1015840

1198863

and 1198631015840

1198863

start to conduct as soon as 11986310158401198862

switches off Sincethe voltage applied to inductor 119871 is the constant 119880

1198892 the

current flowing through 119871 increases linearly with time Incontrast the current flowing through 1198631015840

1198864

decreases linearlywith time (as shown in Figure 7(c)) At the same time whenthe current flowing through 119863

1015840

1198864

decreases to 0 the currentflowing through the snubber inductor is equivalent to theload current and then thewhole commutation process comesto an end 1198631015840

1198864

is blocked 11987910158401198863

and 11986310158401198863

are carrying the loadcurrent and 1198621015840

2

is discharged simultaneously (Figure 7(d))The red lines represent the current path of 1198791015840

1198861

rarr

1198631015840

1198863

1198791015840

1198863

commutation process in S3L inverter during differ-ent periods

It can be seen from the figures that the ratios of the currentflowing through 119878

1015840

1198862

11986310158401198862

11987810158401198863

and 1198631015840

1198863

are limited within alimited range Meanwhile the switching process of 1198781015840

1198863

is softswitching and its power loss is small as well Similarly theratios of currents flowing through1198631015840

1198864

and11986310158401198861

are limited ina certain rangeTherefore a substantial reduction of chargingenergy is realized during the reverse recovery and the powerloss of charging is reduced with it as well

The rest of commutation processes in Table 3 work in asimilar way which is soft switching type and has nothing todo with the influences caused by the amplitude and angle(0∘ sim 360

∘) of load current so it will not be detailedrespectively

4 Simulation and Experiment

Based on the theories and algorithms above the experimentwas conducted on a 21MWexperimental platform (Figures 8and 9) which includes two 21MWmotors and both of themare controlled running under the same conditions by NPCthree-level inverter and S3L inverter respectively to carry outthe comparing experiment of improvement effectiveness

It can be seen by analyzing the waveforms in Figures10 and 11 that the output waveforms of the two three-levelinverters which use the same SPWM modulation algorithmand control parameters are almost consistent in waveformdistortion and harmonic content when the peak value ofoutput phase voltage is 119881dcradic3 Therefore it is consideredthat the S3L inverter has the same output characteristicswith NPC three-level inverter with the same modulationconditions and S3L inverter has a much higher harmoniccontent in output voltage and slightly smaller total distortionrate however its low-order harmonics account for a muchbigger proportion It can be summarized by analyzing Figures12 and 13 that S3L inverter has the same excellent outputwaveforms with NPC3L inverter and its output waveforms ofcurrent are smoothing and approximate sine curve

It can be seen by analyzing Figures 14(a)sim14(f) that theIGBT (1198791015840

1198862

) current surge of S3L inverter is only two-thirds

6 Journal of Electrical and Computer Engineering

Table 2 Commutation processes of S3L inverter

Load current is positive Load current is negativeCommutation Allowed Involved Commutation Allowed Involved1198791015840

1198861

rarr 1198631015840

1198863

11987910158401198863

YES 1198621015840

2

1198631015840

1198861

rarr 1198631015840

1198862

11987910158401198862

YES 1198621015840

2

1198631015840

1198863

11987910158401198863

rarr 1198791015840

1198861

YES 1198621015840

2

1198631015840

1198862

11987910158401198862

rarr 1198631015840

1198861

YES 1198621015840

2

1198631015840

1198863

11987910158401198863

rarr 1198631015840

1198864

YES 1198621015840

1

1198631015840

1198862

11987910158401198862

rarr 1198791015840

1198864

YES 1198621015840

1

1198631015840

1198864

rarr 1198631015840

1198863

11987910158401198863

YES 1198621015840

1

1198791015840

1198864

rarr 1198631015840

1198862

11987910158401198862

YES 1198621015840

1

1198791015840

1198861

rarr 1198631015840

1198864

NO mdash 1198631015840

1198861

rarr 1198791015840

1198864

NO mdash1198631015840

1198864

rarr 1198791015840

1198861

NO mdash 1198791015840

1198864

rarr 1198631015840

1198861

NO mdash

Ud2

Ud2

+

+minus

minus

0

DaT1

DaT2

DaT3

DaT4

L

Load

iL

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(a)

Ud2

Ud2

+

+minus

minus

0

DaT1

DaT2

DaT3

DaT4

L

Load

iL

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(b)

Ud2

Ud2

+

+minus

minus

0

DaT1

DaT2

DaT3

DaT4

L

Load

iL

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(c)

Ud2

Ud2

+

+minus

minus

DaT1

DaT2

DaT3

DaT4

L

Load

iL

0

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(d)

Figure 7 Commutation process of 11987910158401198861

rarr 1198631015840

1198863

11987910158401198863

(a) Before commutation (b) 1199050

le 119905 lt 1199051

period (c) 1199051

le 119905 lt 1199052

period (d) aftercommutation

Table 3 Commutation processes with zero load current

Commutation Allowed Involved1198791015840

1198861

rarr 1198631015840

1198863

11987910158401198863

YES 1198621015840

2

1198631015840

1198863

11987910158401198863

rarr 1198791015840

1198861

YES 1198621015840

2

1198791015840

1198861

rarr 1198631015840

1198864

NO mdash1198631015840

1198861

rarr 1198631015840

1198862

11987910158401198862

YES 1198621015840

1

1198631015840

1198862

11987910158401198862

rarr 1198631015840

1198861

YES 1198621015840

1

1198631015840

1198861

rarr 1198791015840

1198864

YES mdash

of NPC3L inverter (1198791198862) at the switching-on instant the

IGBT (11987910158401198862

) voltage surge of S3L inverter is only half ofNPC3L inverter (119879

1198862) at the switching-off instant overall S3L

Squirrel-cage motor

Double-fed induction

motor

Figure 8 The 21MW dragging platform-motor part

inverters have much lower switching-on and switching-offvoltage and current surges than NPC3L inverters

Journal of Electrical and Computer Engineering 7

NPC3L explosion-proof inverter (1MW) S3L explosion-proof inverter (1MW)

Figure 9 The 21MW dragging platform-inverter part

Time (5msdiv)

CH1

CH2

CH3CH3

(1V

div

)CH

2(1

Vd

iv)

CH1

(1V

div

)

(a) The modulation waveform at119898 = 1154

(b) NPC harmonic content (c) S3L harmonic content

Figure 10 The comparison of phase-voltage harmonic characteristics with119898 = 1154

It is pointed out in the optimized power-loss calculationalgorithm and maternal module concept that it is necessaryto calculate the power losses of other devices except powerdevices on the same bridge to modify the total loss when per-forming the system thermal analysis In this paper themodifi-cation aspects include the power losses of precharge current-limiting resistor balanced resistors absorption capacitanceDC-link capacitors and buffer devices of S3L inverter By theway in any other cases the modification can be calculatedin accordance with the actual conditions These devices areusually fixed in the explosion-proof inverter housing andsome of them are working all the way releasing some power

as constant thermal sources which will elevate the ambienttemperature of the whole cabinet and affect the thermalflow in the cabinet All of this above will finally influencethe power-loss calculation and generate an error betweentheoretical calculation and actual value The conclusion canbe drawn by analyzing Figures 15(a)sim15(d) that the powerloss of IGBT 119879

1198861and its antiparallel diode119863

1198861in NPC three-

level inverters is much bigger than 1198791015840

1198861

and its antiparalleldiode 119863

1015840

1198861

in S3L inverter with different modulations andload impedance angles According to the optimized power-loss calculation algorithm the additional power loss ofeach power device module was calculated and completed

8 Journal of Electrical and Computer Engineering

Time (5msdiv)

CH1

CH2

CH3CH3

(1V

div

)CH

2(1

Vd

iv)

CH1

(1V

div

)(a) The modulation waveform at119898 = 01

(b) NPC harmonic content (c) S3L harmonic content

Figure 11 The comparison of phase-voltage harmonic characteristics with119898 = 01

Time (5msdiv)

CH1 CH2

CH2

(200

Ad

iv)

CH1

(200

Vd

iv)

(a) NPC3L waveforms

Time (5msdiv)

CH1

CH2

CH2

(500

Ad

iv)

CH1

(200

Vd

iv)

(b) S3L waveforms

Figure 12 The comparison of voltage and current in the same IGBT of two inverters

the modification of maternal model power loss in summarythe half-bridge total power loss of S3L inverter maternalmodel is much smaller than NPCrsquos inverter after the mod-ification by 213W on which basis the total power loss ofinverters can be obtained

The calculated power-loss value before modification andaftermodificationwas put into the thermalmodel ofmaternalmodel built by ANSYS ICEPAK respectively as a thermalresource value and the thermal analysis results of twoinverters were presented in Figures 16 and 17 It can be derivedby analyzing the Figures 17(a) and 17(b) that the heat sinktemperature of S3L inverter is about 8∘C lower than that

of NPC three-level inverter running in the same coolingsystems and operating under the same conditions while thesame value in Figures 16(a) and 16(b) before modificationis 3∘C It is easy to find that the substrate temperatureof S3L inverter is 10∘C lower than that of NPC inverterapproximately running in the same cooling systems andoperating under the same conditions by analyzing Figures17(c) and 17(d) But there is a 4∘C decline in Figures 16(c)and 16(d) before modification Overall the analysis showsthat the power devicesrsquo temperature of S3L inverter has a 9∘Cadvantage over NPC inverter under modification and a 35∘Cadvantage without modification

Journal of Electrical and Computer Engineering 9

CH1 AB line voltage (1000Vdiv)

CH2

CH1

CH2 phase A current (500Adiv)

(a) The AB line voltage and phase A current

CH1 AB line voltage (1000Vdiv)

CH1

CH2

CH2 phase A current (500Adiv)

(b) Details of AB line voltage and phase A current

Figure 13 The output voltage and current waveforms of S3L inverter

Time (500nsdiv)

CH1

CH2

CH2

(200

Vd

iv)

CH1

(200

Ad

iv)

(a) IGBT switch-on (NPC)

Time (200nsdiv)

CH1

CH2

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(b) IGBT switch-on (S3L)

Time (200nsdiv)

CH1

CH2

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(c) IGBT switch-off (NPC)

Time (200nsdiv)

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(d) IGBT switch-off (S3L)

Time (200nsdiv)

CH1

(200

Ad

iv)

CH1

(e) NPC diode reverse recovery

Time (200nsdiv)

CH1

CH1

(200

Ad

iv)

(f) S3L diode reverse recovery

Figure 14 The comparative experiments of IGBT and diodersquos characteristics

10 Journal of Electrical and Computer Engineering

500

400

300

200

100

0

Pow

er (W

)

35325

215

105

0

120593 (rad)0

100200

300400

500600

Current (A)

(a) 1198791198861

power loss

400

300

200

100

0

Pow

er (W

)

35325 2

151

050

120593 (rad)0

100 200300

400500

600

Current (A)

(b) 11987910158401198861

power loss

300350

200250

10050

150

0

Powe

r (W

)

35325 2

15105

0

120593 (rad)0 100

200300

400 500600

Current (A)

(c) 1198631198861

power loss

250

300

200

100

50

150

0

Pow

er (W

)

35325

215 1

050

120593 (rad)0 100 200

300400 500

600

Current (A)

(d) 11986310158401198861

power loss

200

150

100

50

0

Pow

er (W

)

35325 2

15105

0 0100

200300

400500

600

Current (A)120593 (rad)

(e) NPC additional power-loss correction

80100120140160180

6040200

Pow

er (W

)

35325 2

15 105

0

120593 (rad)0 100 200

300400

500600

Current (A)

(f) S3L additional power-loss correction

Figure 15The half-bridge power devicesrsquo power-loss calculation of the two inverters based on optimized power-loss algorithm (a)1198791198861

powerloss (b) 1198791015840

1198861

power loss (c) 1198631198861

power loss (d) 11986310158401198861

power loss (e) NPC additional power-loss correction (f) S3L additional power-losscorrection

By observing the experiment results in Figures 18(a)and 18(b) the fact that the heat sink surface temperature ofS3L inverter is about 6∘C lower than that of NPC three-level inverter in average under the same conditions canbe obtained which is in line with the theoretical analysisexpectation by observing the experiment results in Figures18(c) and 18(d) the fact that the IGBT substrate temperatureof S3L inverter is lower than NPC three-level inverter by 11∘Ccan be figured out which is consistent with the theoreticalanalysis results Generally speaking the experimental results

show that the power device temperature of S3L inverter islower than that of NPC three-level inverter by 10∘C approxi-mately which is much closer to the theoretical analysis result9∘C with modification and only has 1∘C error under theseexperiment conditions

In summary the maternal model based on the optimizedpower-loss algorithm has a much higher thermal analysisaccuracy in the improvement process of three-level inverterswhich offers a 1∘C error between theoretical calculationand experiment value in this paper and can be used as

Journal of Electrical and Computer Engineering 11

79828

74736

69643

6455

59458

54365

49272

4418

39087

33994

(a)

7561

70822

66035

61247

5646

51673

46885

42098

37311

32523

(b)

79828

79233

78637

78042

77446

76851

76255

7566

75064

74469

(c)

7561

7502

7443

73841

73251

72662

72072

71483

70893

70303

(d)

Figure 16 The temperature distribution 3D map of maternal thermal models before modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

a tool to support the accurate power-loss calculation andthermal analysis using the S3L inverter based on the softswitching control to improve the NPC three-level invertercan get a good result that S3L inverter has the same excellentoutput characteristic with NPC three-level inverter and hasa great advantage in reducing power loss with a 213Wdecline in each half-bridge and 10∘C decline on power devicetemperatureThe thermal stability of three-level inverters canbe enhanced by this improvement

5 Conclusions

The optimized maternal power-loss thermal models of NPCthree-level inverter and S3L inverter were established based

on the optimized power-loss algorithm and a set of generaloptimized power-loss calculation formulas was derived tomodify the total power loss and figure out the modificationpower-loss values Then these values were considered asthermal sources to analyze the maternal thermal modelsThe three-level inverter can be improved by comparing andanalyzing power-loss modification values and experimentresults Based on this principle and methods the fact thatunder the same conditions the power-lossmodification valueof S3L inverter is smaller than that ofNPC three-level inverterby 213W and has a 9∘C advantage is obtained which is only1∘C smaller than the experiment result Experimental resultsvalidate the proposed theoretical calculation and analysis andprove the effectiveness of the improvement

12 Journal of Electrical and Computer Engineering

61678

58801

55924

53047

5017

47293

44416

41539

38662

35785

(a)

49007

4728

45554

43828

42102

40376

38649

36923

35197

33471

(b)

61678

61374

6107

60766

60462

60159

59855

59551

59247

58943

(c)

49007

48824

48642

4846

48277

48095

47913

47731

47548

47366

(d)

Figure 17 The temperature distribution 3D map of maternal thermal models after modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

Appendix

The parameters of double-fed induction motor used in thisexperiment are as follows

rated power 2100KWrated speed 1000 rpmnumber of pole-pairs 3efficiency at full load 97network voltage 690Vnetwork frequency 50Hzstator current 1662A

rotor current 743Acoupling stator Δ rotor Yrotor open voltage 1710Vinertia 94 kgm2stator maximum short-circuit current 4250Arotor maximum short-circuit current 2900Astator resistance R1 0006013Ωstator leakage reactance X1 0045062Ωrotor resistance (equivalent) R2 0004193Ωrotor leakage reactance (equivalent) X2 02298Ω

Journal of Electrical and Computer Engineering 13

NPC3L

(∘C)

603

56

52

48

44

40

36

32

28

246

(a)

S3L

552

51

48

45

42

39

36

33

30

27

247

(∘C )

(b)

NPC3L

(c)

S3L

(d)

Figure 18The experimental temperature results of IGBTrsquos surface and substrate (a) IGBT surface temperature of NPC3L inverter (b) IGBTsurface temperature of S3L inverter (c) IGBT substrate temperature of NPC3L inverter (d) IGBT substrate temperature of S3L inverter

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors would like to thank 2014 Jiangsu ProvinceNatural Science Foundation (BK20140204) the Research andInnovation Program of Postgraduates in Jiangsu Province(CXZZ13 0930) and the Fundamental Research Funds forthe Central Universities (2012LWB73)

References

[1] S Bernet ldquoRecent developments of high power converters forindustry and traction applicationsrdquo IEEE Transactions on PowerElectronics vol 15 no 6 pp 1102ndash1117 2000

[2] P Mao S-J Xie and Z-G Xu ldquoSwitching transients model andloss analysis of IGBTmodulerdquo Proceedings of the Chinese Societyof Electrical Engineering vol 15 article 007 2010

[3] A D Rajapakse AM Gole and P LWilson ldquoElectromagnetictransients simulation models for accurate representation ofswitching losses and thermal performance in power electronic

systemsrdquo IEEE Transactions on Power Delivery vol 20 no 1 pp319ndash327 2005

[4] J Hu J Li J Zou and J Tan ldquoLosses calculation of IGBTmodule and heat dissipation system design of invertersrdquo Trans-actions of China Electrotechnical Society vol 24 no 3 pp 159ndash163 2009

[5] M H Bierhoff and F W Fuchs ldquoSemiconductor losses involtage source and current source IGBT converters based onanalytical derivationrdquo in Proceedings of the IEEE 35th AnnualPower Electronics Specialists Conference (PESC rsquo04) vol 4 pp2836ndash2842 IEEE June 2004

[6] F Krismer and J W Kolar ldquoAccurate power loss modelderivation of a high-current dual active bridge converter foran automotive applicationrdquo IEEE Transactions on IndustrialElectronics vol 57 no 3 pp 881ndash891 2010

[7] Q Chen Q Wang W Jiang and C Hu ldquoAnalysis of switchinglosses in diode-clamped three-level converterrdquo Transactions ofChina Electrotechnical Society vol 23 no 2 pp 68ndash75 2008

[8] J Wang Q Chen W Jiang and C Hu ldquoAnalysis of conductionlosses in neutral-point-clamped three-level inverterrdquo Transac-tions of China Electrotechnical Society vol 3 p 12 2007

[9] T J Kim DW Kang Y H Lee andD S Hyun ldquoThe analysis ofconduction and switching losses inmulti-level inverter systemrdquoin Proceedings of the IEEE 32nd Annual Power ElectronicsSpecialists Conference (PESC rsquo01) pp 1363ndash1368 June 2001

14 Journal of Electrical and Computer Engineering

[10] S Dieckerhoff S Bernet and D Krug ldquoPower loss-orientedevaluation of high voltage IGBTs and multilevel convertersin transformerless traction applicationsrdquo IEEE Transactions onPower Electronics vol 20 no 6 pp 1328ndash1336 2005

[11] W Jing G Tan and Z Ye ldquoLosses calculation and heat dissipa-tion analysis of high-power three-level convertersrdquoTransactionsof China Electrotechnical Society vol 26 no 2 pp 134ndash140 2011

[12] K Ma Y Yang and F Blaabjerg ldquoTransient modelling ofloss and thermal dynamics in power semiconductor devicesrdquoin Proceedings of the IEEE Energy Conversion Congress andExposition (ECCE rsquo14) pp 5495ndash5501 IEEE 2014

[13] U R Prasanna and A K Rathore ldquoAnalysis design and exper-imental results of a novel soft-switching snubberless current-fed half-bridge front-end converter-based pv inverterrdquo IEEETransactions on Power Electronics vol 28 no 7 pp 3219ndash32302013

[14] X Ruan L Zhou and Y Yan ldquoSoft-switching PWM three-levelconvertersrdquo IEEE Transactions on Power Electronics vol 16 no5 pp 612ndash622 2001

[15] R Li andD Xu ldquoA zero-voltage switching three-phase inverterrdquoIEEE Transactions on Power Electronics vol 29 no 3 pp 1200ndash1210 2014

[16] P Kollensperger R U Lenke S Schroder and R W de Don-cker ldquoDesign of a flexible control platform for soft-switchingmultilevel invertersrdquo IEEE Transactions on Power Electronicsvol 22 no 5 pp 1778ndash1785 2007

[17] G Ortiz H Uemura D Bortis J W Kolar and O ApeldoornldquoModeling of soft-switching losses of IGBTs in high-powerhigh-efficiency dual-active-bridge DCDC convertersrdquo IEEETransactions on Electron Devices vol 60 no 2 pp 587ndash5972013

[18] M W Gekeler ldquoSoft switching three level inverter (S3Linverter)rdquo in Proceedings of the 15th European Conference onPower Electronics and Applications (EPE rsquo13) pp 1ndash10 IEEESeptember 2013

[19] S Munk-Nielsen L N Tutelea and U Jaeger ldquoSimulation withideal switchmodels combinedwithmeasured loss data providesa good estimate of power lossrdquo in Proceedings of the ConferenceRecord of the IEEE Industry Applications Conference vol 5 pp2915ndash2922 Rome Italy October 2000

[20] O S Senturk L Helle S Munk-Nielsen P Rodriguez and RTeodorescu ldquoConverter structure-based power loss and staticthermal modeling of the press-pack IGBT three-level ANPCVSC applied to multi-MW wind turbinesrdquo IEEE Transactionson Industry Applications vol 47 no 6 pp 2505ndash2515 2011

[21] D A B Zambra C Rech F A S Goncalves and J R PinheiroldquoPower losses analysis and cooling system design of threetopologies of multilevel invertersrdquo in Proceedings of the 39thIEEE Annual Power Electronics Specialists Conference (PESCrsquo08) pp 4290ndash4295 IEEE June 2008

[22] W Jing Three-level inverter power loss power devices research[Doctoral dissertation] China University of Mining and Tech-nology Xuzhou China 2011

[23] D Floricau E Floricau and G Gateau ldquoThree-level ActiveNPC converter PWM strategies and loss distributionrdquo inProceedings of the 34th Annual Conference of the IEEE IndustrialElectronics Society (IECON rsquo08) pp 3333ndash3338November 2008

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

2 Journal of Electrical and Computer Engineering

the main reason causing errors between their theoreticalcalculations and experimental results where Dieckerhoff etal [10] considered that the switching power loss of powerdevice has a linear relationship with its withstanding voltagewhile this assumption is approximately valid only in plusmn20range of the test voltage A much accurate losses calculationand heat dissipation method was introduced in [11] but itdid not take all the thermal sources in consideration whichhas an effect on the power devices and thermal analysis Inthe literature [12] the transient modeling of loss and thermaldynamics in power semiconductor devices is analyzed whileit needs to improve the model by considering the peripheralcircuits Several soft switching inverter types and controlmethods are proposed in [13ndash18] where the S3L inverter in[18] has a much more significant effect on the reduction ofpower losses It is the accurate thermal analysis methods ofinverter system that can analyze the inverter temperaturequantitatively providing references for inverter improve-ments [19 20] In the existing loss calculation studies of three-level inverters it lacks a system of theories and methodsto provide theoretical support for the improvements Beforea new three-level topology improving the NPC3L inverterit is necessary to apply a common theoretical calculationand method to anticipate its advantages The S3L inverterproposed in the literature [18] holds the viewpoint that it canreduce the power loss in terms of the NPC3L inverter underthe same conditions but there is no quantitative experimentaltemperature to support it and demonstrate its effectiveness ofimprovement

For the above reasons a general power-loss calculationmethod of three-level inverters was established in this paperbased on the optimized power-loss algorithm in Section 2with which an accurate power-loss calculation and per-formance evaluation approach of three-level inverters wasproposed To improve NPC3L inverter the S3L inverterworking principlewas elaborated in Section 3 andput into thegeneral approach mentioned above What is more accordingto this approach it is expected in Section 4 that the S3Linverter has the same excellent output characteristics withNPC3L inverter and it can reduce power loss by 213Wbringing in a 10∘C decrease in temperature intuitively In thesame section the experiment results support the validity ofthe theoretical prediction Finally Section 5 concludes thispaper

2 General Optimized Power-Loss AlgorithmBased on NPC3L Inverter

The main circuit topology of NPC3L inverter is shown inFigure 1

Each leg has four IGBTs labeled 1198791198941 1198791198942 1198791198943 and 119879

1198944

(where 119894 represents one phase of 119886 119887 and 119888 phases and eachIGBT has one antiparallel diode labeled 119863

1198941 1198631198942 1198631198943 and

1198631198944 resp) and two clamping diodes labeled 119863

1198945and 119863

1198946

[11 21]At present the IGBT device is a power module packaged

by one IGBT and a fast recovery antiparallel diodeTherefore

Udc

P

C1

O

N

C2

Ta1

Ta2

Ta3

Ta4

Da1

Da2

Da3

Da4

Da5

Da6

ab c

UVW

Tb1

Tb2

Tb3

Tb4

Db1

Db2

Db3

Db4

Db5

Db6

Tc1

Tc2

Tc3

Tc4

Dc1

Dc2

Dc3

Dc4

Dc5

Dc6

Figure 1 The main circuit topology of NPC3L inverter

its total power loss is composed of these two parts expressedas follows

119875mod = 119875119879+ 119875119863 (1)

The equivalent structure model of power device and heatsink is shown in Figure 2

It can be seen fromFigure 2 that the wholemodel consistsof four conductive layers and therefore if we consider thepower device and the heat sink as a maternal model the fourconductive layers would be four submodels On this basis thethermal resistance and heat capacity of the four submodelscan be calculated at first respectively and then all the foursubmodelsrsquo thermal resistance and heat capacity constitutethe total thermal resistance and heat capacity of the wholemodel

The calculation formula of thermal resistance 119877th isdescribed as follows

119877th =Δ119879

119875

=Δ119879

119876Δ119905

(2)

where Δ119879 is the temperature increase of submodel 119875 and Δ119905are the heat flow and time period respectivelyThen the totalthermal resistance of maternal model is shown as

119877th-total = 119877th1 + 119877th2 + 119877th3 + 119877th4 (3)

The calculation formula of thermal capacity 119862th is deliv-ered as follows [20 21]

119862th =119876

Δ119879

(4)

The total heat capacity of the maternal model can bewritten as [20 21]

119862th-total = 119862th1 + 119862th2 + 119862th3 + 119862th4 (5)

Thus thematernal model can be replaced alternatively bya fourth-order RC circuit shown in Figure 3

Journal of Electrical and Computer Engineering 3

Chip

Solder

Solder

Copper layer (Al2O3AlN)Ceramics

Copper layer

Board

Heat sink

Submodel 1

Submodel 2

Submodel 3

Submodel 4

Junctiontemperature (Tvj)

Chip-shelltemperature (ΔTjc)

Shell temperature (Tc)Shell heat sink

temperature (ΔTch)

Chip-case thermalresistance Rthjc

Pin-IGBT Pout-IGBT

Heat sinktemperature (Th)

Heat sinkambiance (ΔTha)

Ambienttemperature (Ta)

Case heat sinkRthch

Heat sinkambiance Rthha

Figure 2 Equivalent structure model of power device and heat sink

TvjT Rth1 Rth2 Rth3 Rth4

Cth1 Cth2 Cth3 Cth4PT

Tc

(a)

Cth1 Cth2 Cth3 Cth4PT

TvjTRth1 Rth2 Rth3 Rth4 Tc

(b)

Figure 3 The fourth-order RC thermal resistance equivalent circuits of IGBT (a) Partial network (b) continuous network

TvjT

TvjD

RthjcT

RthjcDRthca

Rthch

RthhaPT PD

Tc

Th

Ta

Figure 4The thermal resistance equivalent circuit of IGBTmodulein steady state

According to the partial network structure of Figure 3(a)the IGBT thermal resistance can be derived as [20 21]

119885thjc119879 =119899

sum

119894=1

119877th119894 (1 minus eminus119905120591119894) (119894 = 1 2 3 4) (6)

where 120591119894is the RC time constant of each layer

The thermal equivalent circuit of IGBT module in steadystate is shown in Figure 4

As a switching device the IGBTrsquos power loss 119875119879

isprimarily composed of conduction loss 119875con119879 and switchingloss 119875sw119879 namely [3ndash5]

119875119879= 119875con119879 + 119875sw119879 (7)

The conduction resistance initial saturation voltage andconduction loss of IGBT can be expressed respectively asfollows [11]

119903119879= 119903119879-25∘C + 119870119903119879 (119879vj119879 minus 25

∘C)

V0119879

= V0119879-25∘C + 119870V0119879 (119879vj119879 minus 25

∘C)

119875con119879 = V0119879119868 + 1199031198791198682

(8)

where 119903119879-25∘C and V

0119879-25∘C are the conduction resistanceand initial saturation voltage of IGBT with the junctiontemperature at 25∘C 119870

119903119879and 119870V0119879 are the initial saturation

voltage and conduction resistance temperature correctionfactor of IGBT 119879vj119879 is the junction temperature of IGBT 119868is the instantaneous current flowing through the IGBT

Combine the three-level working principle and optimizedIGBT power-loss model and the average conduction and

4 Journal of Electrical and Computer Engineering

switching losses formula ofT1in amodulation voltage period

will be as follows [11 22]

119875npcspwmcon1198791

= 1198910

119902

sum

119896=119901

(V01198791

+ 1199031198791119868119871(119896)) 119868119871(119896) 120591 (119896) 119879s

119875npcspwmsw1198791

= 1198910

119902

sum

119896=119901

119864sw1198791 (119868119871 (119896))

(9)

where 1198910is the frequency of modulation voltage 120591(119896) is

the 119896th duty cycle of switching period 119868119871(119896) is the average

load current of the 119896th switching period 119901 and 119902 representthe sampling periodrsquos beginning and end of 119879

1during one

modulation period respectivelyGenerally speaking when the carrier ratio is large

enough the discrete power-loss formula can be transformedinto a continuous integral form and the average conductionlosses and switching losses of 119879

1can be expressed as [22]

119875npcspwmcon1198791

=1

2120587

int

120587

120593

(V01198791

+ 1199031198791119894119871(120572)) 119894119871(120572)119863 (120572) 119889120572

=

119898V01198791

119868119898

4120587

((120587 minus 120593) cos120593 + sin120593)

+

11989811990311987911198682

119898

4120587

(1 +4

3

cos120593 + 1

3

cos 2120593)

119875npcspwmsw1198791

=1

2120587

int

120587

120593

119891sw119864sw1198791 (119894119871 (120572)) 119889120572

=119891sw2120587

(119860 sw1198791198682

119898

1

2

(120587 minus 120593 +1

2

sin 2120593)

+ 119861sw119879119868119898 (1 + cos120593) + 119862sw119879 (120587 minus 120593) )

sdot (119880dc2

119880base)

119863sw119879

(

119879vj1198791119879base

)

119870sw119879

(10)

In accordance with the same calculation principle asT1rsquos the conduction losses and switching losses (or reverse

recovery losses) of119863111987921198632 and119863

5in the same half-bridge

leg will be 119875npcspwmcon1198631 119875npcspwm

rec1198631 119875npcspwmcon1198792

119875npcspwmsw1198792

119875npcspwmcon1198632

119875npcspwmrec1198632 119875npcspwm

con1198635 and 119875npcspwmrec1198635

Some explanatory notes in expressions (8)sim(10) are asfollows

V0119879119909

= V0119879-25∘C + 119870V0119879(119879vj119879119909 minus 25

∘C) denotes the 119909thIBGTrsquos initial saturation voltage119903119879119909

= 119903119879-25∘C + 119870

119903119879(119879vj119879119909 minus 25

∘C) represents the 119909thIGBTrsquos conduction resistance119879vj119879119909 means the 119909th IGBTrsquoS junction temperature

V0119863119909

= V0119863-25∘C + 119870V0119863(119879vj119863119909 minus 25

∘C) indicates the119909th fast recovery diodersquos initial saturation voltage119903119863119909

= 119903119863-25∘C + 119870119903119863(119879vj119863119909 minus 25

∘C) stands for the 119909thfast recovery diodersquos conduction resistance

TvjT Rth1

Cth1

Rth2

Cth2

Rth3

Cth3

Rth4

Cth4PT

Tc

CaddRadd

Figure 5 The equivalent circuit of optimized power-loss model

119879vj119863119909 means the 119909th fast recovery diodersquos junctiontemperature what is more 119909 = 1 2 5 [22]

In accordance with the above step the power losses ofheat sink and IGBT maternal model can be calculated How-ever there may be buffer circuit or something like that in theperiphery around the power devices of different three-leveltopologiesThatmeans part of the power difference (119875in-IGBTminus119875out-IGBT) flowing through the IGBT model is dissipated inthe IGBT model and some other part is consumed by theperipheral circuits namely

119875in-IGBT minus 119875out-IGBT = 119875IGBT + 119875add (11)

Hence it is necessary to optimize the maternal modeland the equivalent circuit of optimized maternal model isshown in Figure 5

According to the equivalent circuit of optimizedmaternalmodel the total power-loss equation of half-bridge leg can bemodified as follows

119875total = 1198751198791+ 1198751198631+ 1198751198792+ 1198751198632+ 119875add (12)

The general power-loss calculation of half-bridge leg ofthree-level inverters using SPWMmodulation algorithm in amodulation period will be the one as follows [22 23]

119875total = 119875npcspwmcon1198791

+ 119875npcspwmsw1198791

+ 119875npcspwmcon1198631

+ 119875npcspwmrec1198631 + 119875

npcspwmcon1198792

+ 119875npcspwmsw1198792

+ 119875npcspwmcon1198632 + 119875

npcspwmrec1198632 + 119875add

(13)

where the SPWM modulation algorithm can be replacedby the actual algorithm but the power-loss calculation hasthe same process based on optimized power-loss algorithmand maternal model and all we should do is to changethe variables consistent with the algorithm we are goingto use In addition during the analysis of maternal mod-ule in Section 4 it is necessary to modify 119875add by con-sidering precharge current-limiting resistor balanced resis-tors absorption capacitance DC-link capacitors and bufferdevices of S3L inverter

3 S3L Inverter Principle

As shown in Figure 6 one full-bridge leg topology of S3Linverter contains four IGBTs (1198791015840

1198861

sim 1198791015840

1198864

) four diodes (11986310158401198861

sim

1198631015840

1198864

) snubber inductor snubber capacitances11986210158401

and11986210158402

andfour snubber diodes 119863

1198861198791sim 1198631198861198794

where the latter fourconstitute the snubber circuit [14]

Journal of Electrical and Computer Engineering 5

Table 1 Switching states of S3L inverter

Switching state + 0 minus

1198801015840

load +119880119889

2 0 minus119880119889

2

Conduction 1198791015840

1198861

or11986310158401198861

1198791015840

1198862

11986310158401198862

or 11987910158401198863

11986310158401198863

1198791015840

1198864

or11986310158401198864

1198791015840

1198861

ON OFF OFF1198791015840

1198862

OFF ON ON1198791015840

1198863

ON ON OFF1198791015840

1198864

OFF OFF ON

Ud2

Ud2

+

minus

P998400

0998400

N998400

DaT1

DaT2

DaT3

DaT4

L

Load

C9984001

C9984002

T998400a1

T998400a2

T998400a3

T998400a4

D998400a1

D998400a2

D998400a3

D998400a4

U998400load

Figure 6 One full-bridge leg topology of S3L inverters

S3L three-level inverter switching state and commutationprocess are shown in Tables 1 and 2

For zero load current commutation process it can beconsidered as three special cases specified in Table 3

Each of these commutation processes is slightly differentand therefore only the 119879

1015840

1198861

rarr 1198631015840

1198863

1198791015840

1198863

was chosen todescribe the working details as an example In order tofacilitate the analysis the load current in the commutationprocess is supposed to be constant substantially and its pathis marked in red

Before the commutation process begins 11987910158401198861

carries thepositive load current 119868Load and 119879

1015840

1198863

is switched on (butdoes not carry current because of diode 1198631015840

1198863

) 11987910158401198862

and 1198791015840

1198864

are switched off The output terminal is connected to thepositive terminal of the input DC voltage The capacitor 1198621015840

1

is discharged the capacitor11986210158402

is charged to minus119880119889 The current

in the snubber inductor is zero (Figure 7(a)) [14]The 1198791015840

1198861

rarr 1198631015840

1198863

1198791015840

1198863

commutation process starts as soonas 11987910158401198861

is switched off when 1198791015840

1198862

is switched off and what ismore 1198791015840

1198863

and 11987910158401198864

remain switched on and off respectively Inaccordance with the different current path and IGBT actionsequences the whole process can be divided into two periods

(1) 1199050le 119905 lt 119905

1Period Two current loops are generated

during this stage One of them is the oscillating current loopconstituted by 1198621015840

2

11987910158401198862

11986310158401198862

119871 1198801198892 and 119863

1198861198794 the other is

the load current loop generated by the load current flowingthrough 1198621015840

2

load midpoint 0 1198801198892 and 119863

1198861198794 As shown in

Figure 7(b) the two current paths overlap each other It isnoteworthy that the current flow decreases rapidly to zeroand the rising slope of the voltage both ends is limited to asmall amplitude so that the power loss is correspondingly

small At this time the switching-off process of 11987910158401198861

is the so-called soft switching

(2) 1199051le 119905 lt 119905

2Period The first period of commutation

process comes to an end when 1198621015840

2

discharges and 1198631015840

1198864

starts conducting At the same time the current flowingthrough inductor 119871 reduces to 0 and what is more 119879

1015840

1198863

and 1198631015840

1198863

start to conduct as soon as 11986310158401198862

switches off Sincethe voltage applied to inductor 119871 is the constant 119880

1198892 the

current flowing through 119871 increases linearly with time Incontrast the current flowing through 1198631015840

1198864

decreases linearlywith time (as shown in Figure 7(c)) At the same time whenthe current flowing through 119863

1015840

1198864

decreases to 0 the currentflowing through the snubber inductor is equivalent to theload current and then thewhole commutation process comesto an end 1198631015840

1198864

is blocked 11987910158401198863

and 11986310158401198863

are carrying the loadcurrent and 1198621015840

2

is discharged simultaneously (Figure 7(d))The red lines represent the current path of 1198791015840

1198861

rarr

1198631015840

1198863

1198791015840

1198863

commutation process in S3L inverter during differ-ent periods

It can be seen from the figures that the ratios of the currentflowing through 119878

1015840

1198862

11986310158401198862

11987810158401198863

and 1198631015840

1198863

are limited within alimited range Meanwhile the switching process of 1198781015840

1198863

is softswitching and its power loss is small as well Similarly theratios of currents flowing through1198631015840

1198864

and11986310158401198861

are limited ina certain rangeTherefore a substantial reduction of chargingenergy is realized during the reverse recovery and the powerloss of charging is reduced with it as well

The rest of commutation processes in Table 3 work in asimilar way which is soft switching type and has nothing todo with the influences caused by the amplitude and angle(0∘ sim 360

∘) of load current so it will not be detailedrespectively

4 Simulation and Experiment

Based on the theories and algorithms above the experimentwas conducted on a 21MWexperimental platform (Figures 8and 9) which includes two 21MWmotors and both of themare controlled running under the same conditions by NPCthree-level inverter and S3L inverter respectively to carry outthe comparing experiment of improvement effectiveness

It can be seen by analyzing the waveforms in Figures10 and 11 that the output waveforms of the two three-levelinverters which use the same SPWM modulation algorithmand control parameters are almost consistent in waveformdistortion and harmonic content when the peak value ofoutput phase voltage is 119881dcradic3 Therefore it is consideredthat the S3L inverter has the same output characteristicswith NPC three-level inverter with the same modulationconditions and S3L inverter has a much higher harmoniccontent in output voltage and slightly smaller total distortionrate however its low-order harmonics account for a muchbigger proportion It can be summarized by analyzing Figures12 and 13 that S3L inverter has the same excellent outputwaveforms with NPC3L inverter and its output waveforms ofcurrent are smoothing and approximate sine curve

It can be seen by analyzing Figures 14(a)sim14(f) that theIGBT (1198791015840

1198862

) current surge of S3L inverter is only two-thirds

6 Journal of Electrical and Computer Engineering

Table 2 Commutation processes of S3L inverter

Load current is positive Load current is negativeCommutation Allowed Involved Commutation Allowed Involved1198791015840

1198861

rarr 1198631015840

1198863

11987910158401198863

YES 1198621015840

2

1198631015840

1198861

rarr 1198631015840

1198862

11987910158401198862

YES 1198621015840

2

1198631015840

1198863

11987910158401198863

rarr 1198791015840

1198861

YES 1198621015840

2

1198631015840

1198862

11987910158401198862

rarr 1198631015840

1198861

YES 1198621015840

2

1198631015840

1198863

11987910158401198863

rarr 1198631015840

1198864

YES 1198621015840

1

1198631015840

1198862

11987910158401198862

rarr 1198791015840

1198864

YES 1198621015840

1

1198631015840

1198864

rarr 1198631015840

1198863

11987910158401198863

YES 1198621015840

1

1198791015840

1198864

rarr 1198631015840

1198862

11987910158401198862

YES 1198621015840

1

1198791015840

1198861

rarr 1198631015840

1198864

NO mdash 1198631015840

1198861

rarr 1198791015840

1198864

NO mdash1198631015840

1198864

rarr 1198791015840

1198861

NO mdash 1198791015840

1198864

rarr 1198631015840

1198861

NO mdash

Ud2

Ud2

+

+minus

minus

0

DaT1

DaT2

DaT3

DaT4

L

Load

iL

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(a)

Ud2

Ud2

+

+minus

minus

0

DaT1

DaT2

DaT3

DaT4

L

Load

iL

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(b)

Ud2

Ud2

+

+minus

minus

0

DaT1

DaT2

DaT3

DaT4

L

Load

iL

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(c)

Ud2

Ud2

+

+minus

minus

DaT1

DaT2

DaT3

DaT4

L

Load

iL

0

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(d)

Figure 7 Commutation process of 11987910158401198861

rarr 1198631015840

1198863

11987910158401198863

(a) Before commutation (b) 1199050

le 119905 lt 1199051

period (c) 1199051

le 119905 lt 1199052

period (d) aftercommutation

Table 3 Commutation processes with zero load current

Commutation Allowed Involved1198791015840

1198861

rarr 1198631015840

1198863

11987910158401198863

YES 1198621015840

2

1198631015840

1198863

11987910158401198863

rarr 1198791015840

1198861

YES 1198621015840

2

1198791015840

1198861

rarr 1198631015840

1198864

NO mdash1198631015840

1198861

rarr 1198631015840

1198862

11987910158401198862

YES 1198621015840

1

1198631015840

1198862

11987910158401198862

rarr 1198631015840

1198861

YES 1198621015840

1

1198631015840

1198861

rarr 1198791015840

1198864

YES mdash

of NPC3L inverter (1198791198862) at the switching-on instant the

IGBT (11987910158401198862

) voltage surge of S3L inverter is only half ofNPC3L inverter (119879

1198862) at the switching-off instant overall S3L

Squirrel-cage motor

Double-fed induction

motor

Figure 8 The 21MW dragging platform-motor part

inverters have much lower switching-on and switching-offvoltage and current surges than NPC3L inverters

Journal of Electrical and Computer Engineering 7

NPC3L explosion-proof inverter (1MW) S3L explosion-proof inverter (1MW)

Figure 9 The 21MW dragging platform-inverter part

Time (5msdiv)

CH1

CH2

CH3CH3

(1V

div

)CH

2(1

Vd

iv)

CH1

(1V

div

)

(a) The modulation waveform at119898 = 1154

(b) NPC harmonic content (c) S3L harmonic content

Figure 10 The comparison of phase-voltage harmonic characteristics with119898 = 1154

It is pointed out in the optimized power-loss calculationalgorithm and maternal module concept that it is necessaryto calculate the power losses of other devices except powerdevices on the same bridge to modify the total loss when per-forming the system thermal analysis In this paper themodifi-cation aspects include the power losses of precharge current-limiting resistor balanced resistors absorption capacitanceDC-link capacitors and buffer devices of S3L inverter By theway in any other cases the modification can be calculatedin accordance with the actual conditions These devices areusually fixed in the explosion-proof inverter housing andsome of them are working all the way releasing some power

as constant thermal sources which will elevate the ambienttemperature of the whole cabinet and affect the thermalflow in the cabinet All of this above will finally influencethe power-loss calculation and generate an error betweentheoretical calculation and actual value The conclusion canbe drawn by analyzing Figures 15(a)sim15(d) that the powerloss of IGBT 119879

1198861and its antiparallel diode119863

1198861in NPC three-

level inverters is much bigger than 1198791015840

1198861

and its antiparalleldiode 119863

1015840

1198861

in S3L inverter with different modulations andload impedance angles According to the optimized power-loss calculation algorithm the additional power loss ofeach power device module was calculated and completed

8 Journal of Electrical and Computer Engineering

Time (5msdiv)

CH1

CH2

CH3CH3

(1V

div

)CH

2(1

Vd

iv)

CH1

(1V

div

)(a) The modulation waveform at119898 = 01

(b) NPC harmonic content (c) S3L harmonic content

Figure 11 The comparison of phase-voltage harmonic characteristics with119898 = 01

Time (5msdiv)

CH1 CH2

CH2

(200

Ad

iv)

CH1

(200

Vd

iv)

(a) NPC3L waveforms

Time (5msdiv)

CH1

CH2

CH2

(500

Ad

iv)

CH1

(200

Vd

iv)

(b) S3L waveforms

Figure 12 The comparison of voltage and current in the same IGBT of two inverters

the modification of maternal model power loss in summarythe half-bridge total power loss of S3L inverter maternalmodel is much smaller than NPCrsquos inverter after the mod-ification by 213W on which basis the total power loss ofinverters can be obtained

The calculated power-loss value before modification andaftermodificationwas put into the thermalmodel ofmaternalmodel built by ANSYS ICEPAK respectively as a thermalresource value and the thermal analysis results of twoinverters were presented in Figures 16 and 17 It can be derivedby analyzing the Figures 17(a) and 17(b) that the heat sinktemperature of S3L inverter is about 8∘C lower than that

of NPC three-level inverter running in the same coolingsystems and operating under the same conditions while thesame value in Figures 16(a) and 16(b) before modificationis 3∘C It is easy to find that the substrate temperatureof S3L inverter is 10∘C lower than that of NPC inverterapproximately running in the same cooling systems andoperating under the same conditions by analyzing Figures17(c) and 17(d) But there is a 4∘C decline in Figures 16(c)and 16(d) before modification Overall the analysis showsthat the power devicesrsquo temperature of S3L inverter has a 9∘Cadvantage over NPC inverter under modification and a 35∘Cadvantage without modification

Journal of Electrical and Computer Engineering 9

CH1 AB line voltage (1000Vdiv)

CH2

CH1

CH2 phase A current (500Adiv)

(a) The AB line voltage and phase A current

CH1 AB line voltage (1000Vdiv)

CH1

CH2

CH2 phase A current (500Adiv)

(b) Details of AB line voltage and phase A current

Figure 13 The output voltage and current waveforms of S3L inverter

Time (500nsdiv)

CH1

CH2

CH2

(200

Vd

iv)

CH1

(200

Ad

iv)

(a) IGBT switch-on (NPC)

Time (200nsdiv)

CH1

CH2

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(b) IGBT switch-on (S3L)

Time (200nsdiv)

CH1

CH2

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(c) IGBT switch-off (NPC)

Time (200nsdiv)

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(d) IGBT switch-off (S3L)

Time (200nsdiv)

CH1

(200

Ad

iv)

CH1

(e) NPC diode reverse recovery

Time (200nsdiv)

CH1

CH1

(200

Ad

iv)

(f) S3L diode reverse recovery

Figure 14 The comparative experiments of IGBT and diodersquos characteristics

10 Journal of Electrical and Computer Engineering

500

400

300

200

100

0

Pow

er (W

)

35325

215

105

0

120593 (rad)0

100200

300400

500600

Current (A)

(a) 1198791198861

power loss

400

300

200

100

0

Pow

er (W

)

35325 2

151

050

120593 (rad)0

100 200300

400500

600

Current (A)

(b) 11987910158401198861

power loss

300350

200250

10050

150

0

Powe

r (W

)

35325 2

15105

0

120593 (rad)0 100

200300

400 500600

Current (A)

(c) 1198631198861

power loss

250

300

200

100

50

150

0

Pow

er (W

)

35325

215 1

050

120593 (rad)0 100 200

300400 500

600

Current (A)

(d) 11986310158401198861

power loss

200

150

100

50

0

Pow

er (W

)

35325 2

15105

0 0100

200300

400500

600

Current (A)120593 (rad)

(e) NPC additional power-loss correction

80100120140160180

6040200

Pow

er (W

)

35325 2

15 105

0

120593 (rad)0 100 200

300400

500600

Current (A)

(f) S3L additional power-loss correction

Figure 15The half-bridge power devicesrsquo power-loss calculation of the two inverters based on optimized power-loss algorithm (a)1198791198861

powerloss (b) 1198791015840

1198861

power loss (c) 1198631198861

power loss (d) 11986310158401198861

power loss (e) NPC additional power-loss correction (f) S3L additional power-losscorrection

By observing the experiment results in Figures 18(a)and 18(b) the fact that the heat sink surface temperature ofS3L inverter is about 6∘C lower than that of NPC three-level inverter in average under the same conditions canbe obtained which is in line with the theoretical analysisexpectation by observing the experiment results in Figures18(c) and 18(d) the fact that the IGBT substrate temperatureof S3L inverter is lower than NPC three-level inverter by 11∘Ccan be figured out which is consistent with the theoreticalanalysis results Generally speaking the experimental results

show that the power device temperature of S3L inverter islower than that of NPC three-level inverter by 10∘C approxi-mately which is much closer to the theoretical analysis result9∘C with modification and only has 1∘C error under theseexperiment conditions

In summary the maternal model based on the optimizedpower-loss algorithm has a much higher thermal analysisaccuracy in the improvement process of three-level inverterswhich offers a 1∘C error between theoretical calculationand experiment value in this paper and can be used as

Journal of Electrical and Computer Engineering 11

79828

74736

69643

6455

59458

54365

49272

4418

39087

33994

(a)

7561

70822

66035

61247

5646

51673

46885

42098

37311

32523

(b)

79828

79233

78637

78042

77446

76851

76255

7566

75064

74469

(c)

7561

7502

7443

73841

73251

72662

72072

71483

70893

70303

(d)

Figure 16 The temperature distribution 3D map of maternal thermal models before modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

a tool to support the accurate power-loss calculation andthermal analysis using the S3L inverter based on the softswitching control to improve the NPC three-level invertercan get a good result that S3L inverter has the same excellentoutput characteristic with NPC three-level inverter and hasa great advantage in reducing power loss with a 213Wdecline in each half-bridge and 10∘C decline on power devicetemperatureThe thermal stability of three-level inverters canbe enhanced by this improvement

5 Conclusions

The optimized maternal power-loss thermal models of NPCthree-level inverter and S3L inverter were established based

on the optimized power-loss algorithm and a set of generaloptimized power-loss calculation formulas was derived tomodify the total power loss and figure out the modificationpower-loss values Then these values were considered asthermal sources to analyze the maternal thermal modelsThe three-level inverter can be improved by comparing andanalyzing power-loss modification values and experimentresults Based on this principle and methods the fact thatunder the same conditions the power-lossmodification valueof S3L inverter is smaller than that ofNPC three-level inverterby 213W and has a 9∘C advantage is obtained which is only1∘C smaller than the experiment result Experimental resultsvalidate the proposed theoretical calculation and analysis andprove the effectiveness of the improvement

12 Journal of Electrical and Computer Engineering

61678

58801

55924

53047

5017

47293

44416

41539

38662

35785

(a)

49007

4728

45554

43828

42102

40376

38649

36923

35197

33471

(b)

61678

61374

6107

60766

60462

60159

59855

59551

59247

58943

(c)

49007

48824

48642

4846

48277

48095

47913

47731

47548

47366

(d)

Figure 17 The temperature distribution 3D map of maternal thermal models after modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

Appendix

The parameters of double-fed induction motor used in thisexperiment are as follows

rated power 2100KWrated speed 1000 rpmnumber of pole-pairs 3efficiency at full load 97network voltage 690Vnetwork frequency 50Hzstator current 1662A

rotor current 743Acoupling stator Δ rotor Yrotor open voltage 1710Vinertia 94 kgm2stator maximum short-circuit current 4250Arotor maximum short-circuit current 2900Astator resistance R1 0006013Ωstator leakage reactance X1 0045062Ωrotor resistance (equivalent) R2 0004193Ωrotor leakage reactance (equivalent) X2 02298Ω

Journal of Electrical and Computer Engineering 13

NPC3L

(∘C)

603

56

52

48

44

40

36

32

28

246

(a)

S3L

552

51

48

45

42

39

36

33

30

27

247

(∘C )

(b)

NPC3L

(c)

S3L

(d)

Figure 18The experimental temperature results of IGBTrsquos surface and substrate (a) IGBT surface temperature of NPC3L inverter (b) IGBTsurface temperature of S3L inverter (c) IGBT substrate temperature of NPC3L inverter (d) IGBT substrate temperature of S3L inverter

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors would like to thank 2014 Jiangsu ProvinceNatural Science Foundation (BK20140204) the Research andInnovation Program of Postgraduates in Jiangsu Province(CXZZ13 0930) and the Fundamental Research Funds forthe Central Universities (2012LWB73)

References

[1] S Bernet ldquoRecent developments of high power converters forindustry and traction applicationsrdquo IEEE Transactions on PowerElectronics vol 15 no 6 pp 1102ndash1117 2000

[2] P Mao S-J Xie and Z-G Xu ldquoSwitching transients model andloss analysis of IGBTmodulerdquo Proceedings of the Chinese Societyof Electrical Engineering vol 15 article 007 2010

[3] A D Rajapakse AM Gole and P LWilson ldquoElectromagnetictransients simulation models for accurate representation ofswitching losses and thermal performance in power electronic

systemsrdquo IEEE Transactions on Power Delivery vol 20 no 1 pp319ndash327 2005

[4] J Hu J Li J Zou and J Tan ldquoLosses calculation of IGBTmodule and heat dissipation system design of invertersrdquo Trans-actions of China Electrotechnical Society vol 24 no 3 pp 159ndash163 2009

[5] M H Bierhoff and F W Fuchs ldquoSemiconductor losses involtage source and current source IGBT converters based onanalytical derivationrdquo in Proceedings of the IEEE 35th AnnualPower Electronics Specialists Conference (PESC rsquo04) vol 4 pp2836ndash2842 IEEE June 2004

[6] F Krismer and J W Kolar ldquoAccurate power loss modelderivation of a high-current dual active bridge converter foran automotive applicationrdquo IEEE Transactions on IndustrialElectronics vol 57 no 3 pp 881ndash891 2010

[7] Q Chen Q Wang W Jiang and C Hu ldquoAnalysis of switchinglosses in diode-clamped three-level converterrdquo Transactions ofChina Electrotechnical Society vol 23 no 2 pp 68ndash75 2008

[8] J Wang Q Chen W Jiang and C Hu ldquoAnalysis of conductionlosses in neutral-point-clamped three-level inverterrdquo Transac-tions of China Electrotechnical Society vol 3 p 12 2007

[9] T J Kim DW Kang Y H Lee andD S Hyun ldquoThe analysis ofconduction and switching losses inmulti-level inverter systemrdquoin Proceedings of the IEEE 32nd Annual Power ElectronicsSpecialists Conference (PESC rsquo01) pp 1363ndash1368 June 2001

14 Journal of Electrical and Computer Engineering

[10] S Dieckerhoff S Bernet and D Krug ldquoPower loss-orientedevaluation of high voltage IGBTs and multilevel convertersin transformerless traction applicationsrdquo IEEE Transactions onPower Electronics vol 20 no 6 pp 1328ndash1336 2005

[11] W Jing G Tan and Z Ye ldquoLosses calculation and heat dissipa-tion analysis of high-power three-level convertersrdquoTransactionsof China Electrotechnical Society vol 26 no 2 pp 134ndash140 2011

[12] K Ma Y Yang and F Blaabjerg ldquoTransient modelling ofloss and thermal dynamics in power semiconductor devicesrdquoin Proceedings of the IEEE Energy Conversion Congress andExposition (ECCE rsquo14) pp 5495ndash5501 IEEE 2014

[13] U R Prasanna and A K Rathore ldquoAnalysis design and exper-imental results of a novel soft-switching snubberless current-fed half-bridge front-end converter-based pv inverterrdquo IEEETransactions on Power Electronics vol 28 no 7 pp 3219ndash32302013

[14] X Ruan L Zhou and Y Yan ldquoSoft-switching PWM three-levelconvertersrdquo IEEE Transactions on Power Electronics vol 16 no5 pp 612ndash622 2001

[15] R Li andD Xu ldquoA zero-voltage switching three-phase inverterrdquoIEEE Transactions on Power Electronics vol 29 no 3 pp 1200ndash1210 2014

[16] P Kollensperger R U Lenke S Schroder and R W de Don-cker ldquoDesign of a flexible control platform for soft-switchingmultilevel invertersrdquo IEEE Transactions on Power Electronicsvol 22 no 5 pp 1778ndash1785 2007

[17] G Ortiz H Uemura D Bortis J W Kolar and O ApeldoornldquoModeling of soft-switching losses of IGBTs in high-powerhigh-efficiency dual-active-bridge DCDC convertersrdquo IEEETransactions on Electron Devices vol 60 no 2 pp 587ndash5972013

[18] M W Gekeler ldquoSoft switching three level inverter (S3Linverter)rdquo in Proceedings of the 15th European Conference onPower Electronics and Applications (EPE rsquo13) pp 1ndash10 IEEESeptember 2013

[19] S Munk-Nielsen L N Tutelea and U Jaeger ldquoSimulation withideal switchmodels combinedwithmeasured loss data providesa good estimate of power lossrdquo in Proceedings of the ConferenceRecord of the IEEE Industry Applications Conference vol 5 pp2915ndash2922 Rome Italy October 2000

[20] O S Senturk L Helle S Munk-Nielsen P Rodriguez and RTeodorescu ldquoConverter structure-based power loss and staticthermal modeling of the press-pack IGBT three-level ANPCVSC applied to multi-MW wind turbinesrdquo IEEE Transactionson Industry Applications vol 47 no 6 pp 2505ndash2515 2011

[21] D A B Zambra C Rech F A S Goncalves and J R PinheiroldquoPower losses analysis and cooling system design of threetopologies of multilevel invertersrdquo in Proceedings of the 39thIEEE Annual Power Electronics Specialists Conference (PESCrsquo08) pp 4290ndash4295 IEEE June 2008

[22] W Jing Three-level inverter power loss power devices research[Doctoral dissertation] China University of Mining and Tech-nology Xuzhou China 2011

[23] D Floricau E Floricau and G Gateau ldquoThree-level ActiveNPC converter PWM strategies and loss distributionrdquo inProceedings of the 34th Annual Conference of the IEEE IndustrialElectronics Society (IECON rsquo08) pp 3333ndash3338November 2008

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

Journal of Electrical and Computer Engineering 3

Chip

Solder

Solder

Copper layer (Al2O3AlN)Ceramics

Copper layer

Board

Heat sink

Submodel 1

Submodel 2

Submodel 3

Submodel 4

Junctiontemperature (Tvj)

Chip-shelltemperature (ΔTjc)

Shell temperature (Tc)Shell heat sink

temperature (ΔTch)

Chip-case thermalresistance Rthjc

Pin-IGBT Pout-IGBT

Heat sinktemperature (Th)

Heat sinkambiance (ΔTha)

Ambienttemperature (Ta)

Case heat sinkRthch

Heat sinkambiance Rthha

Figure 2 Equivalent structure model of power device and heat sink

TvjT Rth1 Rth2 Rth3 Rth4

Cth1 Cth2 Cth3 Cth4PT

Tc

(a)

Cth1 Cth2 Cth3 Cth4PT

TvjTRth1 Rth2 Rth3 Rth4 Tc

(b)

Figure 3 The fourth-order RC thermal resistance equivalent circuits of IGBT (a) Partial network (b) continuous network

TvjT

TvjD

RthjcT

RthjcDRthca

Rthch

RthhaPT PD

Tc

Th

Ta

Figure 4The thermal resistance equivalent circuit of IGBTmodulein steady state

According to the partial network structure of Figure 3(a)the IGBT thermal resistance can be derived as [20 21]

119885thjc119879 =119899

sum

119894=1

119877th119894 (1 minus eminus119905120591119894) (119894 = 1 2 3 4) (6)

where 120591119894is the RC time constant of each layer

The thermal equivalent circuit of IGBT module in steadystate is shown in Figure 4

As a switching device the IGBTrsquos power loss 119875119879

isprimarily composed of conduction loss 119875con119879 and switchingloss 119875sw119879 namely [3ndash5]

119875119879= 119875con119879 + 119875sw119879 (7)

The conduction resistance initial saturation voltage andconduction loss of IGBT can be expressed respectively asfollows [11]

119903119879= 119903119879-25∘C + 119870119903119879 (119879vj119879 minus 25

∘C)

V0119879

= V0119879-25∘C + 119870V0119879 (119879vj119879 minus 25

∘C)

119875con119879 = V0119879119868 + 1199031198791198682

(8)

where 119903119879-25∘C and V

0119879-25∘C are the conduction resistanceand initial saturation voltage of IGBT with the junctiontemperature at 25∘C 119870

119903119879and 119870V0119879 are the initial saturation

voltage and conduction resistance temperature correctionfactor of IGBT 119879vj119879 is the junction temperature of IGBT 119868is the instantaneous current flowing through the IGBT

Combine the three-level working principle and optimizedIGBT power-loss model and the average conduction and

4 Journal of Electrical and Computer Engineering

switching losses formula ofT1in amodulation voltage period

will be as follows [11 22]

119875npcspwmcon1198791

= 1198910

119902

sum

119896=119901

(V01198791

+ 1199031198791119868119871(119896)) 119868119871(119896) 120591 (119896) 119879s

119875npcspwmsw1198791

= 1198910

119902

sum

119896=119901

119864sw1198791 (119868119871 (119896))

(9)

where 1198910is the frequency of modulation voltage 120591(119896) is

the 119896th duty cycle of switching period 119868119871(119896) is the average

load current of the 119896th switching period 119901 and 119902 representthe sampling periodrsquos beginning and end of 119879

1during one

modulation period respectivelyGenerally speaking when the carrier ratio is large

enough the discrete power-loss formula can be transformedinto a continuous integral form and the average conductionlosses and switching losses of 119879

1can be expressed as [22]

119875npcspwmcon1198791

=1

2120587

int

120587

120593

(V01198791

+ 1199031198791119894119871(120572)) 119894119871(120572)119863 (120572) 119889120572

=

119898V01198791

119868119898

4120587

((120587 minus 120593) cos120593 + sin120593)

+

11989811990311987911198682

119898

4120587

(1 +4

3

cos120593 + 1

3

cos 2120593)

119875npcspwmsw1198791

=1

2120587

int

120587

120593

119891sw119864sw1198791 (119894119871 (120572)) 119889120572

=119891sw2120587

(119860 sw1198791198682

119898

1

2

(120587 minus 120593 +1

2

sin 2120593)

+ 119861sw119879119868119898 (1 + cos120593) + 119862sw119879 (120587 minus 120593) )

sdot (119880dc2

119880base)

119863sw119879

(

119879vj1198791119879base

)

119870sw119879

(10)

In accordance with the same calculation principle asT1rsquos the conduction losses and switching losses (or reverse

recovery losses) of119863111987921198632 and119863

5in the same half-bridge

leg will be 119875npcspwmcon1198631 119875npcspwm

rec1198631 119875npcspwmcon1198792

119875npcspwmsw1198792

119875npcspwmcon1198632

119875npcspwmrec1198632 119875npcspwm

con1198635 and 119875npcspwmrec1198635

Some explanatory notes in expressions (8)sim(10) are asfollows

V0119879119909

= V0119879-25∘C + 119870V0119879(119879vj119879119909 minus 25

∘C) denotes the 119909thIBGTrsquos initial saturation voltage119903119879119909

= 119903119879-25∘C + 119870

119903119879(119879vj119879119909 minus 25

∘C) represents the 119909thIGBTrsquos conduction resistance119879vj119879119909 means the 119909th IGBTrsquoS junction temperature

V0119863119909

= V0119863-25∘C + 119870V0119863(119879vj119863119909 minus 25

∘C) indicates the119909th fast recovery diodersquos initial saturation voltage119903119863119909

= 119903119863-25∘C + 119870119903119863(119879vj119863119909 minus 25

∘C) stands for the 119909thfast recovery diodersquos conduction resistance

TvjT Rth1

Cth1

Rth2

Cth2

Rth3

Cth3

Rth4

Cth4PT

Tc

CaddRadd

Figure 5 The equivalent circuit of optimized power-loss model

119879vj119863119909 means the 119909th fast recovery diodersquos junctiontemperature what is more 119909 = 1 2 5 [22]

In accordance with the above step the power losses ofheat sink and IGBT maternal model can be calculated How-ever there may be buffer circuit or something like that in theperiphery around the power devices of different three-leveltopologiesThatmeans part of the power difference (119875in-IGBTminus119875out-IGBT) flowing through the IGBT model is dissipated inthe IGBT model and some other part is consumed by theperipheral circuits namely

119875in-IGBT minus 119875out-IGBT = 119875IGBT + 119875add (11)

Hence it is necessary to optimize the maternal modeland the equivalent circuit of optimized maternal model isshown in Figure 5

According to the equivalent circuit of optimizedmaternalmodel the total power-loss equation of half-bridge leg can bemodified as follows

119875total = 1198751198791+ 1198751198631+ 1198751198792+ 1198751198632+ 119875add (12)

The general power-loss calculation of half-bridge leg ofthree-level inverters using SPWMmodulation algorithm in amodulation period will be the one as follows [22 23]

119875total = 119875npcspwmcon1198791

+ 119875npcspwmsw1198791

+ 119875npcspwmcon1198631

+ 119875npcspwmrec1198631 + 119875

npcspwmcon1198792

+ 119875npcspwmsw1198792

+ 119875npcspwmcon1198632 + 119875

npcspwmrec1198632 + 119875add

(13)

where the SPWM modulation algorithm can be replacedby the actual algorithm but the power-loss calculation hasthe same process based on optimized power-loss algorithmand maternal model and all we should do is to changethe variables consistent with the algorithm we are goingto use In addition during the analysis of maternal mod-ule in Section 4 it is necessary to modify 119875add by con-sidering precharge current-limiting resistor balanced resis-tors absorption capacitance DC-link capacitors and bufferdevices of S3L inverter

3 S3L Inverter Principle

As shown in Figure 6 one full-bridge leg topology of S3Linverter contains four IGBTs (1198791015840

1198861

sim 1198791015840

1198864

) four diodes (11986310158401198861

sim

1198631015840

1198864

) snubber inductor snubber capacitances11986210158401

and11986210158402

andfour snubber diodes 119863

1198861198791sim 1198631198861198794

where the latter fourconstitute the snubber circuit [14]

Journal of Electrical and Computer Engineering 5

Table 1 Switching states of S3L inverter

Switching state + 0 minus

1198801015840

load +119880119889

2 0 minus119880119889

2

Conduction 1198791015840

1198861

or11986310158401198861

1198791015840

1198862

11986310158401198862

or 11987910158401198863

11986310158401198863

1198791015840

1198864

or11986310158401198864

1198791015840

1198861

ON OFF OFF1198791015840

1198862

OFF ON ON1198791015840

1198863

ON ON OFF1198791015840

1198864

OFF OFF ON

Ud2

Ud2

+

minus

P998400

0998400

N998400

DaT1

DaT2

DaT3

DaT4

L

Load

C9984001

C9984002

T998400a1

T998400a2

T998400a3

T998400a4

D998400a1

D998400a2

D998400a3

D998400a4

U998400load

Figure 6 One full-bridge leg topology of S3L inverters

S3L three-level inverter switching state and commutationprocess are shown in Tables 1 and 2

For zero load current commutation process it can beconsidered as three special cases specified in Table 3

Each of these commutation processes is slightly differentand therefore only the 119879

1015840

1198861

rarr 1198631015840

1198863

1198791015840

1198863

was chosen todescribe the working details as an example In order tofacilitate the analysis the load current in the commutationprocess is supposed to be constant substantially and its pathis marked in red

Before the commutation process begins 11987910158401198861

carries thepositive load current 119868Load and 119879

1015840

1198863

is switched on (butdoes not carry current because of diode 1198631015840

1198863

) 11987910158401198862

and 1198791015840

1198864

are switched off The output terminal is connected to thepositive terminal of the input DC voltage The capacitor 1198621015840

1

is discharged the capacitor11986210158402

is charged to minus119880119889 The current

in the snubber inductor is zero (Figure 7(a)) [14]The 1198791015840

1198861

rarr 1198631015840

1198863

1198791015840

1198863

commutation process starts as soonas 11987910158401198861

is switched off when 1198791015840

1198862

is switched off and what ismore 1198791015840

1198863

and 11987910158401198864

remain switched on and off respectively Inaccordance with the different current path and IGBT actionsequences the whole process can be divided into two periods

(1) 1199050le 119905 lt 119905

1Period Two current loops are generated

during this stage One of them is the oscillating current loopconstituted by 1198621015840

2

11987910158401198862

11986310158401198862

119871 1198801198892 and 119863

1198861198794 the other is

the load current loop generated by the load current flowingthrough 1198621015840

2

load midpoint 0 1198801198892 and 119863

1198861198794 As shown in

Figure 7(b) the two current paths overlap each other It isnoteworthy that the current flow decreases rapidly to zeroand the rising slope of the voltage both ends is limited to asmall amplitude so that the power loss is correspondingly

small At this time the switching-off process of 11987910158401198861

is the so-called soft switching

(2) 1199051le 119905 lt 119905

2Period The first period of commutation

process comes to an end when 1198621015840

2

discharges and 1198631015840

1198864

starts conducting At the same time the current flowingthrough inductor 119871 reduces to 0 and what is more 119879

1015840

1198863

and 1198631015840

1198863

start to conduct as soon as 11986310158401198862

switches off Sincethe voltage applied to inductor 119871 is the constant 119880

1198892 the

current flowing through 119871 increases linearly with time Incontrast the current flowing through 1198631015840

1198864

decreases linearlywith time (as shown in Figure 7(c)) At the same time whenthe current flowing through 119863

1015840

1198864

decreases to 0 the currentflowing through the snubber inductor is equivalent to theload current and then thewhole commutation process comesto an end 1198631015840

1198864

is blocked 11987910158401198863

and 11986310158401198863

are carrying the loadcurrent and 1198621015840

2

is discharged simultaneously (Figure 7(d))The red lines represent the current path of 1198791015840

1198861

rarr

1198631015840

1198863

1198791015840

1198863

commutation process in S3L inverter during differ-ent periods

It can be seen from the figures that the ratios of the currentflowing through 119878

1015840

1198862

11986310158401198862

11987810158401198863

and 1198631015840

1198863

are limited within alimited range Meanwhile the switching process of 1198781015840

1198863

is softswitching and its power loss is small as well Similarly theratios of currents flowing through1198631015840

1198864

and11986310158401198861

are limited ina certain rangeTherefore a substantial reduction of chargingenergy is realized during the reverse recovery and the powerloss of charging is reduced with it as well

The rest of commutation processes in Table 3 work in asimilar way which is soft switching type and has nothing todo with the influences caused by the amplitude and angle(0∘ sim 360

∘) of load current so it will not be detailedrespectively

4 Simulation and Experiment

Based on the theories and algorithms above the experimentwas conducted on a 21MWexperimental platform (Figures 8and 9) which includes two 21MWmotors and both of themare controlled running under the same conditions by NPCthree-level inverter and S3L inverter respectively to carry outthe comparing experiment of improvement effectiveness

It can be seen by analyzing the waveforms in Figures10 and 11 that the output waveforms of the two three-levelinverters which use the same SPWM modulation algorithmand control parameters are almost consistent in waveformdistortion and harmonic content when the peak value ofoutput phase voltage is 119881dcradic3 Therefore it is consideredthat the S3L inverter has the same output characteristicswith NPC three-level inverter with the same modulationconditions and S3L inverter has a much higher harmoniccontent in output voltage and slightly smaller total distortionrate however its low-order harmonics account for a muchbigger proportion It can be summarized by analyzing Figures12 and 13 that S3L inverter has the same excellent outputwaveforms with NPC3L inverter and its output waveforms ofcurrent are smoothing and approximate sine curve

It can be seen by analyzing Figures 14(a)sim14(f) that theIGBT (1198791015840

1198862

) current surge of S3L inverter is only two-thirds

6 Journal of Electrical and Computer Engineering

Table 2 Commutation processes of S3L inverter

Load current is positive Load current is negativeCommutation Allowed Involved Commutation Allowed Involved1198791015840

1198861

rarr 1198631015840

1198863

11987910158401198863

YES 1198621015840

2

1198631015840

1198861

rarr 1198631015840

1198862

11987910158401198862

YES 1198621015840

2

1198631015840

1198863

11987910158401198863

rarr 1198791015840

1198861

YES 1198621015840

2

1198631015840

1198862

11987910158401198862

rarr 1198631015840

1198861

YES 1198621015840

2

1198631015840

1198863

11987910158401198863

rarr 1198631015840

1198864

YES 1198621015840

1

1198631015840

1198862

11987910158401198862

rarr 1198791015840

1198864

YES 1198621015840

1

1198631015840

1198864

rarr 1198631015840

1198863

11987910158401198863

YES 1198621015840

1

1198791015840

1198864

rarr 1198631015840

1198862

11987910158401198862

YES 1198621015840

1

1198791015840

1198861

rarr 1198631015840

1198864

NO mdash 1198631015840

1198861

rarr 1198791015840

1198864

NO mdash1198631015840

1198864

rarr 1198791015840

1198861

NO mdash 1198791015840

1198864

rarr 1198631015840

1198861

NO mdash

Ud2

Ud2

+

+minus

minus

0

DaT1

DaT2

DaT3

DaT4

L

Load

iL

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(a)

Ud2

Ud2

+

+minus

minus

0

DaT1

DaT2

DaT3

DaT4

L

Load

iL

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(b)

Ud2

Ud2

+

+minus

minus

0

DaT1

DaT2

DaT3

DaT4

L

Load

iL

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(c)

Ud2

Ud2

+

+minus

minus

DaT1

DaT2

DaT3

DaT4

L

Load

iL

0

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(d)

Figure 7 Commutation process of 11987910158401198861

rarr 1198631015840

1198863

11987910158401198863

(a) Before commutation (b) 1199050

le 119905 lt 1199051

period (c) 1199051

le 119905 lt 1199052

period (d) aftercommutation

Table 3 Commutation processes with zero load current

Commutation Allowed Involved1198791015840

1198861

rarr 1198631015840

1198863

11987910158401198863

YES 1198621015840

2

1198631015840

1198863

11987910158401198863

rarr 1198791015840

1198861

YES 1198621015840

2

1198791015840

1198861

rarr 1198631015840

1198864

NO mdash1198631015840

1198861

rarr 1198631015840

1198862

11987910158401198862

YES 1198621015840

1

1198631015840

1198862

11987910158401198862

rarr 1198631015840

1198861

YES 1198621015840

1

1198631015840

1198861

rarr 1198791015840

1198864

YES mdash

of NPC3L inverter (1198791198862) at the switching-on instant the

IGBT (11987910158401198862

) voltage surge of S3L inverter is only half ofNPC3L inverter (119879

1198862) at the switching-off instant overall S3L

Squirrel-cage motor

Double-fed induction

motor

Figure 8 The 21MW dragging platform-motor part

inverters have much lower switching-on and switching-offvoltage and current surges than NPC3L inverters

Journal of Electrical and Computer Engineering 7

NPC3L explosion-proof inverter (1MW) S3L explosion-proof inverter (1MW)

Figure 9 The 21MW dragging platform-inverter part

Time (5msdiv)

CH1

CH2

CH3CH3

(1V

div

)CH

2(1

Vd

iv)

CH1

(1V

div

)

(a) The modulation waveform at119898 = 1154

(b) NPC harmonic content (c) S3L harmonic content

Figure 10 The comparison of phase-voltage harmonic characteristics with119898 = 1154

It is pointed out in the optimized power-loss calculationalgorithm and maternal module concept that it is necessaryto calculate the power losses of other devices except powerdevices on the same bridge to modify the total loss when per-forming the system thermal analysis In this paper themodifi-cation aspects include the power losses of precharge current-limiting resistor balanced resistors absorption capacitanceDC-link capacitors and buffer devices of S3L inverter By theway in any other cases the modification can be calculatedin accordance with the actual conditions These devices areusually fixed in the explosion-proof inverter housing andsome of them are working all the way releasing some power

as constant thermal sources which will elevate the ambienttemperature of the whole cabinet and affect the thermalflow in the cabinet All of this above will finally influencethe power-loss calculation and generate an error betweentheoretical calculation and actual value The conclusion canbe drawn by analyzing Figures 15(a)sim15(d) that the powerloss of IGBT 119879

1198861and its antiparallel diode119863

1198861in NPC three-

level inverters is much bigger than 1198791015840

1198861

and its antiparalleldiode 119863

1015840

1198861

in S3L inverter with different modulations andload impedance angles According to the optimized power-loss calculation algorithm the additional power loss ofeach power device module was calculated and completed

8 Journal of Electrical and Computer Engineering

Time (5msdiv)

CH1

CH2

CH3CH3

(1V

div

)CH

2(1

Vd

iv)

CH1

(1V

div

)(a) The modulation waveform at119898 = 01

(b) NPC harmonic content (c) S3L harmonic content

Figure 11 The comparison of phase-voltage harmonic characteristics with119898 = 01

Time (5msdiv)

CH1 CH2

CH2

(200

Ad

iv)

CH1

(200

Vd

iv)

(a) NPC3L waveforms

Time (5msdiv)

CH1

CH2

CH2

(500

Ad

iv)

CH1

(200

Vd

iv)

(b) S3L waveforms

Figure 12 The comparison of voltage and current in the same IGBT of two inverters

the modification of maternal model power loss in summarythe half-bridge total power loss of S3L inverter maternalmodel is much smaller than NPCrsquos inverter after the mod-ification by 213W on which basis the total power loss ofinverters can be obtained

The calculated power-loss value before modification andaftermodificationwas put into the thermalmodel ofmaternalmodel built by ANSYS ICEPAK respectively as a thermalresource value and the thermal analysis results of twoinverters were presented in Figures 16 and 17 It can be derivedby analyzing the Figures 17(a) and 17(b) that the heat sinktemperature of S3L inverter is about 8∘C lower than that

of NPC three-level inverter running in the same coolingsystems and operating under the same conditions while thesame value in Figures 16(a) and 16(b) before modificationis 3∘C It is easy to find that the substrate temperatureof S3L inverter is 10∘C lower than that of NPC inverterapproximately running in the same cooling systems andoperating under the same conditions by analyzing Figures17(c) and 17(d) But there is a 4∘C decline in Figures 16(c)and 16(d) before modification Overall the analysis showsthat the power devicesrsquo temperature of S3L inverter has a 9∘Cadvantage over NPC inverter under modification and a 35∘Cadvantage without modification

Journal of Electrical and Computer Engineering 9

CH1 AB line voltage (1000Vdiv)

CH2

CH1

CH2 phase A current (500Adiv)

(a) The AB line voltage and phase A current

CH1 AB line voltage (1000Vdiv)

CH1

CH2

CH2 phase A current (500Adiv)

(b) Details of AB line voltage and phase A current

Figure 13 The output voltage and current waveforms of S3L inverter

Time (500nsdiv)

CH1

CH2

CH2

(200

Vd

iv)

CH1

(200

Ad

iv)

(a) IGBT switch-on (NPC)

Time (200nsdiv)

CH1

CH2

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(b) IGBT switch-on (S3L)

Time (200nsdiv)

CH1

CH2

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(c) IGBT switch-off (NPC)

Time (200nsdiv)

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(d) IGBT switch-off (S3L)

Time (200nsdiv)

CH1

(200

Ad

iv)

CH1

(e) NPC diode reverse recovery

Time (200nsdiv)

CH1

CH1

(200

Ad

iv)

(f) S3L diode reverse recovery

Figure 14 The comparative experiments of IGBT and diodersquos characteristics

10 Journal of Electrical and Computer Engineering

500

400

300

200

100

0

Pow

er (W

)

35325

215

105

0

120593 (rad)0

100200

300400

500600

Current (A)

(a) 1198791198861

power loss

400

300

200

100

0

Pow

er (W

)

35325 2

151

050

120593 (rad)0

100 200300

400500

600

Current (A)

(b) 11987910158401198861

power loss

300350

200250

10050

150

0

Powe

r (W

)

35325 2

15105

0

120593 (rad)0 100

200300

400 500600

Current (A)

(c) 1198631198861

power loss

250

300

200

100

50

150

0

Pow

er (W

)

35325

215 1

050

120593 (rad)0 100 200

300400 500

600

Current (A)

(d) 11986310158401198861

power loss

200

150

100

50

0

Pow

er (W

)

35325 2

15105

0 0100

200300

400500

600

Current (A)120593 (rad)

(e) NPC additional power-loss correction

80100120140160180

6040200

Pow

er (W

)

35325 2

15 105

0

120593 (rad)0 100 200

300400

500600

Current (A)

(f) S3L additional power-loss correction

Figure 15The half-bridge power devicesrsquo power-loss calculation of the two inverters based on optimized power-loss algorithm (a)1198791198861

powerloss (b) 1198791015840

1198861

power loss (c) 1198631198861

power loss (d) 11986310158401198861

power loss (e) NPC additional power-loss correction (f) S3L additional power-losscorrection

By observing the experiment results in Figures 18(a)and 18(b) the fact that the heat sink surface temperature ofS3L inverter is about 6∘C lower than that of NPC three-level inverter in average under the same conditions canbe obtained which is in line with the theoretical analysisexpectation by observing the experiment results in Figures18(c) and 18(d) the fact that the IGBT substrate temperatureof S3L inverter is lower than NPC three-level inverter by 11∘Ccan be figured out which is consistent with the theoreticalanalysis results Generally speaking the experimental results

show that the power device temperature of S3L inverter islower than that of NPC three-level inverter by 10∘C approxi-mately which is much closer to the theoretical analysis result9∘C with modification and only has 1∘C error under theseexperiment conditions

In summary the maternal model based on the optimizedpower-loss algorithm has a much higher thermal analysisaccuracy in the improvement process of three-level inverterswhich offers a 1∘C error between theoretical calculationand experiment value in this paper and can be used as

Journal of Electrical and Computer Engineering 11

79828

74736

69643

6455

59458

54365

49272

4418

39087

33994

(a)

7561

70822

66035

61247

5646

51673

46885

42098

37311

32523

(b)

79828

79233

78637

78042

77446

76851

76255

7566

75064

74469

(c)

7561

7502

7443

73841

73251

72662

72072

71483

70893

70303

(d)

Figure 16 The temperature distribution 3D map of maternal thermal models before modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

a tool to support the accurate power-loss calculation andthermal analysis using the S3L inverter based on the softswitching control to improve the NPC three-level invertercan get a good result that S3L inverter has the same excellentoutput characteristic with NPC three-level inverter and hasa great advantage in reducing power loss with a 213Wdecline in each half-bridge and 10∘C decline on power devicetemperatureThe thermal stability of three-level inverters canbe enhanced by this improvement

5 Conclusions

The optimized maternal power-loss thermal models of NPCthree-level inverter and S3L inverter were established based

on the optimized power-loss algorithm and a set of generaloptimized power-loss calculation formulas was derived tomodify the total power loss and figure out the modificationpower-loss values Then these values were considered asthermal sources to analyze the maternal thermal modelsThe three-level inverter can be improved by comparing andanalyzing power-loss modification values and experimentresults Based on this principle and methods the fact thatunder the same conditions the power-lossmodification valueof S3L inverter is smaller than that ofNPC three-level inverterby 213W and has a 9∘C advantage is obtained which is only1∘C smaller than the experiment result Experimental resultsvalidate the proposed theoretical calculation and analysis andprove the effectiveness of the improvement

12 Journal of Electrical and Computer Engineering

61678

58801

55924

53047

5017

47293

44416

41539

38662

35785

(a)

49007

4728

45554

43828

42102

40376

38649

36923

35197

33471

(b)

61678

61374

6107

60766

60462

60159

59855

59551

59247

58943

(c)

49007

48824

48642

4846

48277

48095

47913

47731

47548

47366

(d)

Figure 17 The temperature distribution 3D map of maternal thermal models after modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

Appendix

The parameters of double-fed induction motor used in thisexperiment are as follows

rated power 2100KWrated speed 1000 rpmnumber of pole-pairs 3efficiency at full load 97network voltage 690Vnetwork frequency 50Hzstator current 1662A

rotor current 743Acoupling stator Δ rotor Yrotor open voltage 1710Vinertia 94 kgm2stator maximum short-circuit current 4250Arotor maximum short-circuit current 2900Astator resistance R1 0006013Ωstator leakage reactance X1 0045062Ωrotor resistance (equivalent) R2 0004193Ωrotor leakage reactance (equivalent) X2 02298Ω

Journal of Electrical and Computer Engineering 13

NPC3L

(∘C)

603

56

52

48

44

40

36

32

28

246

(a)

S3L

552

51

48

45

42

39

36

33

30

27

247

(∘C )

(b)

NPC3L

(c)

S3L

(d)

Figure 18The experimental temperature results of IGBTrsquos surface and substrate (a) IGBT surface temperature of NPC3L inverter (b) IGBTsurface temperature of S3L inverter (c) IGBT substrate temperature of NPC3L inverter (d) IGBT substrate temperature of S3L inverter

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors would like to thank 2014 Jiangsu ProvinceNatural Science Foundation (BK20140204) the Research andInnovation Program of Postgraduates in Jiangsu Province(CXZZ13 0930) and the Fundamental Research Funds forthe Central Universities (2012LWB73)

References

[1] S Bernet ldquoRecent developments of high power converters forindustry and traction applicationsrdquo IEEE Transactions on PowerElectronics vol 15 no 6 pp 1102ndash1117 2000

[2] P Mao S-J Xie and Z-G Xu ldquoSwitching transients model andloss analysis of IGBTmodulerdquo Proceedings of the Chinese Societyof Electrical Engineering vol 15 article 007 2010

[3] A D Rajapakse AM Gole and P LWilson ldquoElectromagnetictransients simulation models for accurate representation ofswitching losses and thermal performance in power electronic

systemsrdquo IEEE Transactions on Power Delivery vol 20 no 1 pp319ndash327 2005

[4] J Hu J Li J Zou and J Tan ldquoLosses calculation of IGBTmodule and heat dissipation system design of invertersrdquo Trans-actions of China Electrotechnical Society vol 24 no 3 pp 159ndash163 2009

[5] M H Bierhoff and F W Fuchs ldquoSemiconductor losses involtage source and current source IGBT converters based onanalytical derivationrdquo in Proceedings of the IEEE 35th AnnualPower Electronics Specialists Conference (PESC rsquo04) vol 4 pp2836ndash2842 IEEE June 2004

[6] F Krismer and J W Kolar ldquoAccurate power loss modelderivation of a high-current dual active bridge converter foran automotive applicationrdquo IEEE Transactions on IndustrialElectronics vol 57 no 3 pp 881ndash891 2010

[7] Q Chen Q Wang W Jiang and C Hu ldquoAnalysis of switchinglosses in diode-clamped three-level converterrdquo Transactions ofChina Electrotechnical Society vol 23 no 2 pp 68ndash75 2008

[8] J Wang Q Chen W Jiang and C Hu ldquoAnalysis of conductionlosses in neutral-point-clamped three-level inverterrdquo Transac-tions of China Electrotechnical Society vol 3 p 12 2007

[9] T J Kim DW Kang Y H Lee andD S Hyun ldquoThe analysis ofconduction and switching losses inmulti-level inverter systemrdquoin Proceedings of the IEEE 32nd Annual Power ElectronicsSpecialists Conference (PESC rsquo01) pp 1363ndash1368 June 2001

14 Journal of Electrical and Computer Engineering

[10] S Dieckerhoff S Bernet and D Krug ldquoPower loss-orientedevaluation of high voltage IGBTs and multilevel convertersin transformerless traction applicationsrdquo IEEE Transactions onPower Electronics vol 20 no 6 pp 1328ndash1336 2005

[11] W Jing G Tan and Z Ye ldquoLosses calculation and heat dissipa-tion analysis of high-power three-level convertersrdquoTransactionsof China Electrotechnical Society vol 26 no 2 pp 134ndash140 2011

[12] K Ma Y Yang and F Blaabjerg ldquoTransient modelling ofloss and thermal dynamics in power semiconductor devicesrdquoin Proceedings of the IEEE Energy Conversion Congress andExposition (ECCE rsquo14) pp 5495ndash5501 IEEE 2014

[13] U R Prasanna and A K Rathore ldquoAnalysis design and exper-imental results of a novel soft-switching snubberless current-fed half-bridge front-end converter-based pv inverterrdquo IEEETransactions on Power Electronics vol 28 no 7 pp 3219ndash32302013

[14] X Ruan L Zhou and Y Yan ldquoSoft-switching PWM three-levelconvertersrdquo IEEE Transactions on Power Electronics vol 16 no5 pp 612ndash622 2001

[15] R Li andD Xu ldquoA zero-voltage switching three-phase inverterrdquoIEEE Transactions on Power Electronics vol 29 no 3 pp 1200ndash1210 2014

[16] P Kollensperger R U Lenke S Schroder and R W de Don-cker ldquoDesign of a flexible control platform for soft-switchingmultilevel invertersrdquo IEEE Transactions on Power Electronicsvol 22 no 5 pp 1778ndash1785 2007

[17] G Ortiz H Uemura D Bortis J W Kolar and O ApeldoornldquoModeling of soft-switching losses of IGBTs in high-powerhigh-efficiency dual-active-bridge DCDC convertersrdquo IEEETransactions on Electron Devices vol 60 no 2 pp 587ndash5972013

[18] M W Gekeler ldquoSoft switching three level inverter (S3Linverter)rdquo in Proceedings of the 15th European Conference onPower Electronics and Applications (EPE rsquo13) pp 1ndash10 IEEESeptember 2013

[19] S Munk-Nielsen L N Tutelea and U Jaeger ldquoSimulation withideal switchmodels combinedwithmeasured loss data providesa good estimate of power lossrdquo in Proceedings of the ConferenceRecord of the IEEE Industry Applications Conference vol 5 pp2915ndash2922 Rome Italy October 2000

[20] O S Senturk L Helle S Munk-Nielsen P Rodriguez and RTeodorescu ldquoConverter structure-based power loss and staticthermal modeling of the press-pack IGBT three-level ANPCVSC applied to multi-MW wind turbinesrdquo IEEE Transactionson Industry Applications vol 47 no 6 pp 2505ndash2515 2011

[21] D A B Zambra C Rech F A S Goncalves and J R PinheiroldquoPower losses analysis and cooling system design of threetopologies of multilevel invertersrdquo in Proceedings of the 39thIEEE Annual Power Electronics Specialists Conference (PESCrsquo08) pp 4290ndash4295 IEEE June 2008

[22] W Jing Three-level inverter power loss power devices research[Doctoral dissertation] China University of Mining and Tech-nology Xuzhou China 2011

[23] D Floricau E Floricau and G Gateau ldquoThree-level ActiveNPC converter PWM strategies and loss distributionrdquo inProceedings of the 34th Annual Conference of the IEEE IndustrialElectronics Society (IECON rsquo08) pp 3333ndash3338November 2008

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

4 Journal of Electrical and Computer Engineering

switching losses formula ofT1in amodulation voltage period

will be as follows [11 22]

119875npcspwmcon1198791

= 1198910

119902

sum

119896=119901

(V01198791

+ 1199031198791119868119871(119896)) 119868119871(119896) 120591 (119896) 119879s

119875npcspwmsw1198791

= 1198910

119902

sum

119896=119901

119864sw1198791 (119868119871 (119896))

(9)

where 1198910is the frequency of modulation voltage 120591(119896) is

the 119896th duty cycle of switching period 119868119871(119896) is the average

load current of the 119896th switching period 119901 and 119902 representthe sampling periodrsquos beginning and end of 119879

1during one

modulation period respectivelyGenerally speaking when the carrier ratio is large

enough the discrete power-loss formula can be transformedinto a continuous integral form and the average conductionlosses and switching losses of 119879

1can be expressed as [22]

119875npcspwmcon1198791

=1

2120587

int

120587

120593

(V01198791

+ 1199031198791119894119871(120572)) 119894119871(120572)119863 (120572) 119889120572

=

119898V01198791

119868119898

4120587

((120587 minus 120593) cos120593 + sin120593)

+

11989811990311987911198682

119898

4120587

(1 +4

3

cos120593 + 1

3

cos 2120593)

119875npcspwmsw1198791

=1

2120587

int

120587

120593

119891sw119864sw1198791 (119894119871 (120572)) 119889120572

=119891sw2120587

(119860 sw1198791198682

119898

1

2

(120587 minus 120593 +1

2

sin 2120593)

+ 119861sw119879119868119898 (1 + cos120593) + 119862sw119879 (120587 minus 120593) )

sdot (119880dc2

119880base)

119863sw119879

(

119879vj1198791119879base

)

119870sw119879

(10)

In accordance with the same calculation principle asT1rsquos the conduction losses and switching losses (or reverse

recovery losses) of119863111987921198632 and119863

5in the same half-bridge

leg will be 119875npcspwmcon1198631 119875npcspwm

rec1198631 119875npcspwmcon1198792

119875npcspwmsw1198792

119875npcspwmcon1198632

119875npcspwmrec1198632 119875npcspwm

con1198635 and 119875npcspwmrec1198635

Some explanatory notes in expressions (8)sim(10) are asfollows

V0119879119909

= V0119879-25∘C + 119870V0119879(119879vj119879119909 minus 25

∘C) denotes the 119909thIBGTrsquos initial saturation voltage119903119879119909

= 119903119879-25∘C + 119870

119903119879(119879vj119879119909 minus 25

∘C) represents the 119909thIGBTrsquos conduction resistance119879vj119879119909 means the 119909th IGBTrsquoS junction temperature

V0119863119909

= V0119863-25∘C + 119870V0119863(119879vj119863119909 minus 25

∘C) indicates the119909th fast recovery diodersquos initial saturation voltage119903119863119909

= 119903119863-25∘C + 119870119903119863(119879vj119863119909 minus 25

∘C) stands for the 119909thfast recovery diodersquos conduction resistance

TvjT Rth1

Cth1

Rth2

Cth2

Rth3

Cth3

Rth4

Cth4PT

Tc

CaddRadd

Figure 5 The equivalent circuit of optimized power-loss model

119879vj119863119909 means the 119909th fast recovery diodersquos junctiontemperature what is more 119909 = 1 2 5 [22]

In accordance with the above step the power losses ofheat sink and IGBT maternal model can be calculated How-ever there may be buffer circuit or something like that in theperiphery around the power devices of different three-leveltopologiesThatmeans part of the power difference (119875in-IGBTminus119875out-IGBT) flowing through the IGBT model is dissipated inthe IGBT model and some other part is consumed by theperipheral circuits namely

119875in-IGBT minus 119875out-IGBT = 119875IGBT + 119875add (11)

Hence it is necessary to optimize the maternal modeland the equivalent circuit of optimized maternal model isshown in Figure 5

According to the equivalent circuit of optimizedmaternalmodel the total power-loss equation of half-bridge leg can bemodified as follows

119875total = 1198751198791+ 1198751198631+ 1198751198792+ 1198751198632+ 119875add (12)

The general power-loss calculation of half-bridge leg ofthree-level inverters using SPWMmodulation algorithm in amodulation period will be the one as follows [22 23]

119875total = 119875npcspwmcon1198791

+ 119875npcspwmsw1198791

+ 119875npcspwmcon1198631

+ 119875npcspwmrec1198631 + 119875

npcspwmcon1198792

+ 119875npcspwmsw1198792

+ 119875npcspwmcon1198632 + 119875

npcspwmrec1198632 + 119875add

(13)

where the SPWM modulation algorithm can be replacedby the actual algorithm but the power-loss calculation hasthe same process based on optimized power-loss algorithmand maternal model and all we should do is to changethe variables consistent with the algorithm we are goingto use In addition during the analysis of maternal mod-ule in Section 4 it is necessary to modify 119875add by con-sidering precharge current-limiting resistor balanced resis-tors absorption capacitance DC-link capacitors and bufferdevices of S3L inverter

3 S3L Inverter Principle

As shown in Figure 6 one full-bridge leg topology of S3Linverter contains four IGBTs (1198791015840

1198861

sim 1198791015840

1198864

) four diodes (11986310158401198861

sim

1198631015840

1198864

) snubber inductor snubber capacitances11986210158401

and11986210158402

andfour snubber diodes 119863

1198861198791sim 1198631198861198794

where the latter fourconstitute the snubber circuit [14]

Journal of Electrical and Computer Engineering 5

Table 1 Switching states of S3L inverter

Switching state + 0 minus

1198801015840

load +119880119889

2 0 minus119880119889

2

Conduction 1198791015840

1198861

or11986310158401198861

1198791015840

1198862

11986310158401198862

or 11987910158401198863

11986310158401198863

1198791015840

1198864

or11986310158401198864

1198791015840

1198861

ON OFF OFF1198791015840

1198862

OFF ON ON1198791015840

1198863

ON ON OFF1198791015840

1198864

OFF OFF ON

Ud2

Ud2

+

minus

P998400

0998400

N998400

DaT1

DaT2

DaT3

DaT4

L

Load

C9984001

C9984002

T998400a1

T998400a2

T998400a3

T998400a4

D998400a1

D998400a2

D998400a3

D998400a4

U998400load

Figure 6 One full-bridge leg topology of S3L inverters

S3L three-level inverter switching state and commutationprocess are shown in Tables 1 and 2

For zero load current commutation process it can beconsidered as three special cases specified in Table 3

Each of these commutation processes is slightly differentand therefore only the 119879

1015840

1198861

rarr 1198631015840

1198863

1198791015840

1198863

was chosen todescribe the working details as an example In order tofacilitate the analysis the load current in the commutationprocess is supposed to be constant substantially and its pathis marked in red

Before the commutation process begins 11987910158401198861

carries thepositive load current 119868Load and 119879

1015840

1198863

is switched on (butdoes not carry current because of diode 1198631015840

1198863

) 11987910158401198862

and 1198791015840

1198864

are switched off The output terminal is connected to thepositive terminal of the input DC voltage The capacitor 1198621015840

1

is discharged the capacitor11986210158402

is charged to minus119880119889 The current

in the snubber inductor is zero (Figure 7(a)) [14]The 1198791015840

1198861

rarr 1198631015840

1198863

1198791015840

1198863

commutation process starts as soonas 11987910158401198861

is switched off when 1198791015840

1198862

is switched off and what ismore 1198791015840

1198863

and 11987910158401198864

remain switched on and off respectively Inaccordance with the different current path and IGBT actionsequences the whole process can be divided into two periods

(1) 1199050le 119905 lt 119905

1Period Two current loops are generated

during this stage One of them is the oscillating current loopconstituted by 1198621015840

2

11987910158401198862

11986310158401198862

119871 1198801198892 and 119863

1198861198794 the other is

the load current loop generated by the load current flowingthrough 1198621015840

2

load midpoint 0 1198801198892 and 119863

1198861198794 As shown in

Figure 7(b) the two current paths overlap each other It isnoteworthy that the current flow decreases rapidly to zeroand the rising slope of the voltage both ends is limited to asmall amplitude so that the power loss is correspondingly

small At this time the switching-off process of 11987910158401198861

is the so-called soft switching

(2) 1199051le 119905 lt 119905

2Period The first period of commutation

process comes to an end when 1198621015840

2

discharges and 1198631015840

1198864

starts conducting At the same time the current flowingthrough inductor 119871 reduces to 0 and what is more 119879

1015840

1198863

and 1198631015840

1198863

start to conduct as soon as 11986310158401198862

switches off Sincethe voltage applied to inductor 119871 is the constant 119880

1198892 the

current flowing through 119871 increases linearly with time Incontrast the current flowing through 1198631015840

1198864

decreases linearlywith time (as shown in Figure 7(c)) At the same time whenthe current flowing through 119863

1015840

1198864

decreases to 0 the currentflowing through the snubber inductor is equivalent to theload current and then thewhole commutation process comesto an end 1198631015840

1198864

is blocked 11987910158401198863

and 11986310158401198863

are carrying the loadcurrent and 1198621015840

2

is discharged simultaneously (Figure 7(d))The red lines represent the current path of 1198791015840

1198861

rarr

1198631015840

1198863

1198791015840

1198863

commutation process in S3L inverter during differ-ent periods

It can be seen from the figures that the ratios of the currentflowing through 119878

1015840

1198862

11986310158401198862

11987810158401198863

and 1198631015840

1198863

are limited within alimited range Meanwhile the switching process of 1198781015840

1198863

is softswitching and its power loss is small as well Similarly theratios of currents flowing through1198631015840

1198864

and11986310158401198861

are limited ina certain rangeTherefore a substantial reduction of chargingenergy is realized during the reverse recovery and the powerloss of charging is reduced with it as well

The rest of commutation processes in Table 3 work in asimilar way which is soft switching type and has nothing todo with the influences caused by the amplitude and angle(0∘ sim 360

∘) of load current so it will not be detailedrespectively

4 Simulation and Experiment

Based on the theories and algorithms above the experimentwas conducted on a 21MWexperimental platform (Figures 8and 9) which includes two 21MWmotors and both of themare controlled running under the same conditions by NPCthree-level inverter and S3L inverter respectively to carry outthe comparing experiment of improvement effectiveness

It can be seen by analyzing the waveforms in Figures10 and 11 that the output waveforms of the two three-levelinverters which use the same SPWM modulation algorithmand control parameters are almost consistent in waveformdistortion and harmonic content when the peak value ofoutput phase voltage is 119881dcradic3 Therefore it is consideredthat the S3L inverter has the same output characteristicswith NPC three-level inverter with the same modulationconditions and S3L inverter has a much higher harmoniccontent in output voltage and slightly smaller total distortionrate however its low-order harmonics account for a muchbigger proportion It can be summarized by analyzing Figures12 and 13 that S3L inverter has the same excellent outputwaveforms with NPC3L inverter and its output waveforms ofcurrent are smoothing and approximate sine curve

It can be seen by analyzing Figures 14(a)sim14(f) that theIGBT (1198791015840

1198862

) current surge of S3L inverter is only two-thirds

6 Journal of Electrical and Computer Engineering

Table 2 Commutation processes of S3L inverter

Load current is positive Load current is negativeCommutation Allowed Involved Commutation Allowed Involved1198791015840

1198861

rarr 1198631015840

1198863

11987910158401198863

YES 1198621015840

2

1198631015840

1198861

rarr 1198631015840

1198862

11987910158401198862

YES 1198621015840

2

1198631015840

1198863

11987910158401198863

rarr 1198791015840

1198861

YES 1198621015840

2

1198631015840

1198862

11987910158401198862

rarr 1198631015840

1198861

YES 1198621015840

2

1198631015840

1198863

11987910158401198863

rarr 1198631015840

1198864

YES 1198621015840

1

1198631015840

1198862

11987910158401198862

rarr 1198791015840

1198864

YES 1198621015840

1

1198631015840

1198864

rarr 1198631015840

1198863

11987910158401198863

YES 1198621015840

1

1198791015840

1198864

rarr 1198631015840

1198862

11987910158401198862

YES 1198621015840

1

1198791015840

1198861

rarr 1198631015840

1198864

NO mdash 1198631015840

1198861

rarr 1198791015840

1198864

NO mdash1198631015840

1198864

rarr 1198791015840

1198861

NO mdash 1198791015840

1198864

rarr 1198631015840

1198861

NO mdash

Ud2

Ud2

+

+minus

minus

0

DaT1

DaT2

DaT3

DaT4

L

Load

iL

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(a)

Ud2

Ud2

+

+minus

minus

0

DaT1

DaT2

DaT3

DaT4

L

Load

iL

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(b)

Ud2

Ud2

+

+minus

minus

0

DaT1

DaT2

DaT3

DaT4

L

Load

iL

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(c)

Ud2

Ud2

+

+minus

minus

DaT1

DaT2

DaT3

DaT4

L

Load

iL

0

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(d)

Figure 7 Commutation process of 11987910158401198861

rarr 1198631015840

1198863

11987910158401198863

(a) Before commutation (b) 1199050

le 119905 lt 1199051

period (c) 1199051

le 119905 lt 1199052

period (d) aftercommutation

Table 3 Commutation processes with zero load current

Commutation Allowed Involved1198791015840

1198861

rarr 1198631015840

1198863

11987910158401198863

YES 1198621015840

2

1198631015840

1198863

11987910158401198863

rarr 1198791015840

1198861

YES 1198621015840

2

1198791015840

1198861

rarr 1198631015840

1198864

NO mdash1198631015840

1198861

rarr 1198631015840

1198862

11987910158401198862

YES 1198621015840

1

1198631015840

1198862

11987910158401198862

rarr 1198631015840

1198861

YES 1198621015840

1

1198631015840

1198861

rarr 1198791015840

1198864

YES mdash

of NPC3L inverter (1198791198862) at the switching-on instant the

IGBT (11987910158401198862

) voltage surge of S3L inverter is only half ofNPC3L inverter (119879

1198862) at the switching-off instant overall S3L

Squirrel-cage motor

Double-fed induction

motor

Figure 8 The 21MW dragging platform-motor part

inverters have much lower switching-on and switching-offvoltage and current surges than NPC3L inverters

Journal of Electrical and Computer Engineering 7

NPC3L explosion-proof inverter (1MW) S3L explosion-proof inverter (1MW)

Figure 9 The 21MW dragging platform-inverter part

Time (5msdiv)

CH1

CH2

CH3CH3

(1V

div

)CH

2(1

Vd

iv)

CH1

(1V

div

)

(a) The modulation waveform at119898 = 1154

(b) NPC harmonic content (c) S3L harmonic content

Figure 10 The comparison of phase-voltage harmonic characteristics with119898 = 1154

It is pointed out in the optimized power-loss calculationalgorithm and maternal module concept that it is necessaryto calculate the power losses of other devices except powerdevices on the same bridge to modify the total loss when per-forming the system thermal analysis In this paper themodifi-cation aspects include the power losses of precharge current-limiting resistor balanced resistors absorption capacitanceDC-link capacitors and buffer devices of S3L inverter By theway in any other cases the modification can be calculatedin accordance with the actual conditions These devices areusually fixed in the explosion-proof inverter housing andsome of them are working all the way releasing some power

as constant thermal sources which will elevate the ambienttemperature of the whole cabinet and affect the thermalflow in the cabinet All of this above will finally influencethe power-loss calculation and generate an error betweentheoretical calculation and actual value The conclusion canbe drawn by analyzing Figures 15(a)sim15(d) that the powerloss of IGBT 119879

1198861and its antiparallel diode119863

1198861in NPC three-

level inverters is much bigger than 1198791015840

1198861

and its antiparalleldiode 119863

1015840

1198861

in S3L inverter with different modulations andload impedance angles According to the optimized power-loss calculation algorithm the additional power loss ofeach power device module was calculated and completed

8 Journal of Electrical and Computer Engineering

Time (5msdiv)

CH1

CH2

CH3CH3

(1V

div

)CH

2(1

Vd

iv)

CH1

(1V

div

)(a) The modulation waveform at119898 = 01

(b) NPC harmonic content (c) S3L harmonic content

Figure 11 The comparison of phase-voltage harmonic characteristics with119898 = 01

Time (5msdiv)

CH1 CH2

CH2

(200

Ad

iv)

CH1

(200

Vd

iv)

(a) NPC3L waveforms

Time (5msdiv)

CH1

CH2

CH2

(500

Ad

iv)

CH1

(200

Vd

iv)

(b) S3L waveforms

Figure 12 The comparison of voltage and current in the same IGBT of two inverters

the modification of maternal model power loss in summarythe half-bridge total power loss of S3L inverter maternalmodel is much smaller than NPCrsquos inverter after the mod-ification by 213W on which basis the total power loss ofinverters can be obtained

The calculated power-loss value before modification andaftermodificationwas put into the thermalmodel ofmaternalmodel built by ANSYS ICEPAK respectively as a thermalresource value and the thermal analysis results of twoinverters were presented in Figures 16 and 17 It can be derivedby analyzing the Figures 17(a) and 17(b) that the heat sinktemperature of S3L inverter is about 8∘C lower than that

of NPC three-level inverter running in the same coolingsystems and operating under the same conditions while thesame value in Figures 16(a) and 16(b) before modificationis 3∘C It is easy to find that the substrate temperatureof S3L inverter is 10∘C lower than that of NPC inverterapproximately running in the same cooling systems andoperating under the same conditions by analyzing Figures17(c) and 17(d) But there is a 4∘C decline in Figures 16(c)and 16(d) before modification Overall the analysis showsthat the power devicesrsquo temperature of S3L inverter has a 9∘Cadvantage over NPC inverter under modification and a 35∘Cadvantage without modification

Journal of Electrical and Computer Engineering 9

CH1 AB line voltage (1000Vdiv)

CH2

CH1

CH2 phase A current (500Adiv)

(a) The AB line voltage and phase A current

CH1 AB line voltage (1000Vdiv)

CH1

CH2

CH2 phase A current (500Adiv)

(b) Details of AB line voltage and phase A current

Figure 13 The output voltage and current waveforms of S3L inverter

Time (500nsdiv)

CH1

CH2

CH2

(200

Vd

iv)

CH1

(200

Ad

iv)

(a) IGBT switch-on (NPC)

Time (200nsdiv)

CH1

CH2

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(b) IGBT switch-on (S3L)

Time (200nsdiv)

CH1

CH2

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(c) IGBT switch-off (NPC)

Time (200nsdiv)

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(d) IGBT switch-off (S3L)

Time (200nsdiv)

CH1

(200

Ad

iv)

CH1

(e) NPC diode reverse recovery

Time (200nsdiv)

CH1

CH1

(200

Ad

iv)

(f) S3L diode reverse recovery

Figure 14 The comparative experiments of IGBT and diodersquos characteristics

10 Journal of Electrical and Computer Engineering

500

400

300

200

100

0

Pow

er (W

)

35325

215

105

0

120593 (rad)0

100200

300400

500600

Current (A)

(a) 1198791198861

power loss

400

300

200

100

0

Pow

er (W

)

35325 2

151

050

120593 (rad)0

100 200300

400500

600

Current (A)

(b) 11987910158401198861

power loss

300350

200250

10050

150

0

Powe

r (W

)

35325 2

15105

0

120593 (rad)0 100

200300

400 500600

Current (A)

(c) 1198631198861

power loss

250

300

200

100

50

150

0

Pow

er (W

)

35325

215 1

050

120593 (rad)0 100 200

300400 500

600

Current (A)

(d) 11986310158401198861

power loss

200

150

100

50

0

Pow

er (W

)

35325 2

15105

0 0100

200300

400500

600

Current (A)120593 (rad)

(e) NPC additional power-loss correction

80100120140160180

6040200

Pow

er (W

)

35325 2

15 105

0

120593 (rad)0 100 200

300400

500600

Current (A)

(f) S3L additional power-loss correction

Figure 15The half-bridge power devicesrsquo power-loss calculation of the two inverters based on optimized power-loss algorithm (a)1198791198861

powerloss (b) 1198791015840

1198861

power loss (c) 1198631198861

power loss (d) 11986310158401198861

power loss (e) NPC additional power-loss correction (f) S3L additional power-losscorrection

By observing the experiment results in Figures 18(a)and 18(b) the fact that the heat sink surface temperature ofS3L inverter is about 6∘C lower than that of NPC three-level inverter in average under the same conditions canbe obtained which is in line with the theoretical analysisexpectation by observing the experiment results in Figures18(c) and 18(d) the fact that the IGBT substrate temperatureof S3L inverter is lower than NPC three-level inverter by 11∘Ccan be figured out which is consistent with the theoreticalanalysis results Generally speaking the experimental results

show that the power device temperature of S3L inverter islower than that of NPC three-level inverter by 10∘C approxi-mately which is much closer to the theoretical analysis result9∘C with modification and only has 1∘C error under theseexperiment conditions

In summary the maternal model based on the optimizedpower-loss algorithm has a much higher thermal analysisaccuracy in the improvement process of three-level inverterswhich offers a 1∘C error between theoretical calculationand experiment value in this paper and can be used as

Journal of Electrical and Computer Engineering 11

79828

74736

69643

6455

59458

54365

49272

4418

39087

33994

(a)

7561

70822

66035

61247

5646

51673

46885

42098

37311

32523

(b)

79828

79233

78637

78042

77446

76851

76255

7566

75064

74469

(c)

7561

7502

7443

73841

73251

72662

72072

71483

70893

70303

(d)

Figure 16 The temperature distribution 3D map of maternal thermal models before modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

a tool to support the accurate power-loss calculation andthermal analysis using the S3L inverter based on the softswitching control to improve the NPC three-level invertercan get a good result that S3L inverter has the same excellentoutput characteristic with NPC three-level inverter and hasa great advantage in reducing power loss with a 213Wdecline in each half-bridge and 10∘C decline on power devicetemperatureThe thermal stability of three-level inverters canbe enhanced by this improvement

5 Conclusions

The optimized maternal power-loss thermal models of NPCthree-level inverter and S3L inverter were established based

on the optimized power-loss algorithm and a set of generaloptimized power-loss calculation formulas was derived tomodify the total power loss and figure out the modificationpower-loss values Then these values were considered asthermal sources to analyze the maternal thermal modelsThe three-level inverter can be improved by comparing andanalyzing power-loss modification values and experimentresults Based on this principle and methods the fact thatunder the same conditions the power-lossmodification valueof S3L inverter is smaller than that ofNPC three-level inverterby 213W and has a 9∘C advantage is obtained which is only1∘C smaller than the experiment result Experimental resultsvalidate the proposed theoretical calculation and analysis andprove the effectiveness of the improvement

12 Journal of Electrical and Computer Engineering

61678

58801

55924

53047

5017

47293

44416

41539

38662

35785

(a)

49007

4728

45554

43828

42102

40376

38649

36923

35197

33471

(b)

61678

61374

6107

60766

60462

60159

59855

59551

59247

58943

(c)

49007

48824

48642

4846

48277

48095

47913

47731

47548

47366

(d)

Figure 17 The temperature distribution 3D map of maternal thermal models after modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

Appendix

The parameters of double-fed induction motor used in thisexperiment are as follows

rated power 2100KWrated speed 1000 rpmnumber of pole-pairs 3efficiency at full load 97network voltage 690Vnetwork frequency 50Hzstator current 1662A

rotor current 743Acoupling stator Δ rotor Yrotor open voltage 1710Vinertia 94 kgm2stator maximum short-circuit current 4250Arotor maximum short-circuit current 2900Astator resistance R1 0006013Ωstator leakage reactance X1 0045062Ωrotor resistance (equivalent) R2 0004193Ωrotor leakage reactance (equivalent) X2 02298Ω

Journal of Electrical and Computer Engineering 13

NPC3L

(∘C)

603

56

52

48

44

40

36

32

28

246

(a)

S3L

552

51

48

45

42

39

36

33

30

27

247

(∘C )

(b)

NPC3L

(c)

S3L

(d)

Figure 18The experimental temperature results of IGBTrsquos surface and substrate (a) IGBT surface temperature of NPC3L inverter (b) IGBTsurface temperature of S3L inverter (c) IGBT substrate temperature of NPC3L inverter (d) IGBT substrate temperature of S3L inverter

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors would like to thank 2014 Jiangsu ProvinceNatural Science Foundation (BK20140204) the Research andInnovation Program of Postgraduates in Jiangsu Province(CXZZ13 0930) and the Fundamental Research Funds forthe Central Universities (2012LWB73)

References

[1] S Bernet ldquoRecent developments of high power converters forindustry and traction applicationsrdquo IEEE Transactions on PowerElectronics vol 15 no 6 pp 1102ndash1117 2000

[2] P Mao S-J Xie and Z-G Xu ldquoSwitching transients model andloss analysis of IGBTmodulerdquo Proceedings of the Chinese Societyof Electrical Engineering vol 15 article 007 2010

[3] A D Rajapakse AM Gole and P LWilson ldquoElectromagnetictransients simulation models for accurate representation ofswitching losses and thermal performance in power electronic

systemsrdquo IEEE Transactions on Power Delivery vol 20 no 1 pp319ndash327 2005

[4] J Hu J Li J Zou and J Tan ldquoLosses calculation of IGBTmodule and heat dissipation system design of invertersrdquo Trans-actions of China Electrotechnical Society vol 24 no 3 pp 159ndash163 2009

[5] M H Bierhoff and F W Fuchs ldquoSemiconductor losses involtage source and current source IGBT converters based onanalytical derivationrdquo in Proceedings of the IEEE 35th AnnualPower Electronics Specialists Conference (PESC rsquo04) vol 4 pp2836ndash2842 IEEE June 2004

[6] F Krismer and J W Kolar ldquoAccurate power loss modelderivation of a high-current dual active bridge converter foran automotive applicationrdquo IEEE Transactions on IndustrialElectronics vol 57 no 3 pp 881ndash891 2010

[7] Q Chen Q Wang W Jiang and C Hu ldquoAnalysis of switchinglosses in diode-clamped three-level converterrdquo Transactions ofChina Electrotechnical Society vol 23 no 2 pp 68ndash75 2008

[8] J Wang Q Chen W Jiang and C Hu ldquoAnalysis of conductionlosses in neutral-point-clamped three-level inverterrdquo Transac-tions of China Electrotechnical Society vol 3 p 12 2007

[9] T J Kim DW Kang Y H Lee andD S Hyun ldquoThe analysis ofconduction and switching losses inmulti-level inverter systemrdquoin Proceedings of the IEEE 32nd Annual Power ElectronicsSpecialists Conference (PESC rsquo01) pp 1363ndash1368 June 2001

14 Journal of Electrical and Computer Engineering

[10] S Dieckerhoff S Bernet and D Krug ldquoPower loss-orientedevaluation of high voltage IGBTs and multilevel convertersin transformerless traction applicationsrdquo IEEE Transactions onPower Electronics vol 20 no 6 pp 1328ndash1336 2005

[11] W Jing G Tan and Z Ye ldquoLosses calculation and heat dissipa-tion analysis of high-power three-level convertersrdquoTransactionsof China Electrotechnical Society vol 26 no 2 pp 134ndash140 2011

[12] K Ma Y Yang and F Blaabjerg ldquoTransient modelling ofloss and thermal dynamics in power semiconductor devicesrdquoin Proceedings of the IEEE Energy Conversion Congress andExposition (ECCE rsquo14) pp 5495ndash5501 IEEE 2014

[13] U R Prasanna and A K Rathore ldquoAnalysis design and exper-imental results of a novel soft-switching snubberless current-fed half-bridge front-end converter-based pv inverterrdquo IEEETransactions on Power Electronics vol 28 no 7 pp 3219ndash32302013

[14] X Ruan L Zhou and Y Yan ldquoSoft-switching PWM three-levelconvertersrdquo IEEE Transactions on Power Electronics vol 16 no5 pp 612ndash622 2001

[15] R Li andD Xu ldquoA zero-voltage switching three-phase inverterrdquoIEEE Transactions on Power Electronics vol 29 no 3 pp 1200ndash1210 2014

[16] P Kollensperger R U Lenke S Schroder and R W de Don-cker ldquoDesign of a flexible control platform for soft-switchingmultilevel invertersrdquo IEEE Transactions on Power Electronicsvol 22 no 5 pp 1778ndash1785 2007

[17] G Ortiz H Uemura D Bortis J W Kolar and O ApeldoornldquoModeling of soft-switching losses of IGBTs in high-powerhigh-efficiency dual-active-bridge DCDC convertersrdquo IEEETransactions on Electron Devices vol 60 no 2 pp 587ndash5972013

[18] M W Gekeler ldquoSoft switching three level inverter (S3Linverter)rdquo in Proceedings of the 15th European Conference onPower Electronics and Applications (EPE rsquo13) pp 1ndash10 IEEESeptember 2013

[19] S Munk-Nielsen L N Tutelea and U Jaeger ldquoSimulation withideal switchmodels combinedwithmeasured loss data providesa good estimate of power lossrdquo in Proceedings of the ConferenceRecord of the IEEE Industry Applications Conference vol 5 pp2915ndash2922 Rome Italy October 2000

[20] O S Senturk L Helle S Munk-Nielsen P Rodriguez and RTeodorescu ldquoConverter structure-based power loss and staticthermal modeling of the press-pack IGBT three-level ANPCVSC applied to multi-MW wind turbinesrdquo IEEE Transactionson Industry Applications vol 47 no 6 pp 2505ndash2515 2011

[21] D A B Zambra C Rech F A S Goncalves and J R PinheiroldquoPower losses analysis and cooling system design of threetopologies of multilevel invertersrdquo in Proceedings of the 39thIEEE Annual Power Electronics Specialists Conference (PESCrsquo08) pp 4290ndash4295 IEEE June 2008

[22] W Jing Three-level inverter power loss power devices research[Doctoral dissertation] China University of Mining and Tech-nology Xuzhou China 2011

[23] D Floricau E Floricau and G Gateau ldquoThree-level ActiveNPC converter PWM strategies and loss distributionrdquo inProceedings of the 34th Annual Conference of the IEEE IndustrialElectronics Society (IECON rsquo08) pp 3333ndash3338November 2008

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

Journal of Electrical and Computer Engineering 5

Table 1 Switching states of S3L inverter

Switching state + 0 minus

1198801015840

load +119880119889

2 0 minus119880119889

2

Conduction 1198791015840

1198861

or11986310158401198861

1198791015840

1198862

11986310158401198862

or 11987910158401198863

11986310158401198863

1198791015840

1198864

or11986310158401198864

1198791015840

1198861

ON OFF OFF1198791015840

1198862

OFF ON ON1198791015840

1198863

ON ON OFF1198791015840

1198864

OFF OFF ON

Ud2

Ud2

+

minus

P998400

0998400

N998400

DaT1

DaT2

DaT3

DaT4

L

Load

C9984001

C9984002

T998400a1

T998400a2

T998400a3

T998400a4

D998400a1

D998400a2

D998400a3

D998400a4

U998400load

Figure 6 One full-bridge leg topology of S3L inverters

S3L three-level inverter switching state and commutationprocess are shown in Tables 1 and 2

For zero load current commutation process it can beconsidered as three special cases specified in Table 3

Each of these commutation processes is slightly differentand therefore only the 119879

1015840

1198861

rarr 1198631015840

1198863

1198791015840

1198863

was chosen todescribe the working details as an example In order tofacilitate the analysis the load current in the commutationprocess is supposed to be constant substantially and its pathis marked in red

Before the commutation process begins 11987910158401198861

carries thepositive load current 119868Load and 119879

1015840

1198863

is switched on (butdoes not carry current because of diode 1198631015840

1198863

) 11987910158401198862

and 1198791015840

1198864

are switched off The output terminal is connected to thepositive terminal of the input DC voltage The capacitor 1198621015840

1

is discharged the capacitor11986210158402

is charged to minus119880119889 The current

in the snubber inductor is zero (Figure 7(a)) [14]The 1198791015840

1198861

rarr 1198631015840

1198863

1198791015840

1198863

commutation process starts as soonas 11987910158401198861

is switched off when 1198791015840

1198862

is switched off and what ismore 1198791015840

1198863

and 11987910158401198864

remain switched on and off respectively Inaccordance with the different current path and IGBT actionsequences the whole process can be divided into two periods

(1) 1199050le 119905 lt 119905

1Period Two current loops are generated

during this stage One of them is the oscillating current loopconstituted by 1198621015840

2

11987910158401198862

11986310158401198862

119871 1198801198892 and 119863

1198861198794 the other is

the load current loop generated by the load current flowingthrough 1198621015840

2

load midpoint 0 1198801198892 and 119863

1198861198794 As shown in

Figure 7(b) the two current paths overlap each other It isnoteworthy that the current flow decreases rapidly to zeroand the rising slope of the voltage both ends is limited to asmall amplitude so that the power loss is correspondingly

small At this time the switching-off process of 11987910158401198861

is the so-called soft switching

(2) 1199051le 119905 lt 119905

2Period The first period of commutation

process comes to an end when 1198621015840

2

discharges and 1198631015840

1198864

starts conducting At the same time the current flowingthrough inductor 119871 reduces to 0 and what is more 119879

1015840

1198863

and 1198631015840

1198863

start to conduct as soon as 11986310158401198862

switches off Sincethe voltage applied to inductor 119871 is the constant 119880

1198892 the

current flowing through 119871 increases linearly with time Incontrast the current flowing through 1198631015840

1198864

decreases linearlywith time (as shown in Figure 7(c)) At the same time whenthe current flowing through 119863

1015840

1198864

decreases to 0 the currentflowing through the snubber inductor is equivalent to theload current and then thewhole commutation process comesto an end 1198631015840

1198864

is blocked 11987910158401198863

and 11986310158401198863

are carrying the loadcurrent and 1198621015840

2

is discharged simultaneously (Figure 7(d))The red lines represent the current path of 1198791015840

1198861

rarr

1198631015840

1198863

1198791015840

1198863

commutation process in S3L inverter during differ-ent periods

It can be seen from the figures that the ratios of the currentflowing through 119878

1015840

1198862

11986310158401198862

11987810158401198863

and 1198631015840

1198863

are limited within alimited range Meanwhile the switching process of 1198781015840

1198863

is softswitching and its power loss is small as well Similarly theratios of currents flowing through1198631015840

1198864

and11986310158401198861

are limited ina certain rangeTherefore a substantial reduction of chargingenergy is realized during the reverse recovery and the powerloss of charging is reduced with it as well

The rest of commutation processes in Table 3 work in asimilar way which is soft switching type and has nothing todo with the influences caused by the amplitude and angle(0∘ sim 360

∘) of load current so it will not be detailedrespectively

4 Simulation and Experiment

Based on the theories and algorithms above the experimentwas conducted on a 21MWexperimental platform (Figures 8and 9) which includes two 21MWmotors and both of themare controlled running under the same conditions by NPCthree-level inverter and S3L inverter respectively to carry outthe comparing experiment of improvement effectiveness

It can be seen by analyzing the waveforms in Figures10 and 11 that the output waveforms of the two three-levelinverters which use the same SPWM modulation algorithmand control parameters are almost consistent in waveformdistortion and harmonic content when the peak value ofoutput phase voltage is 119881dcradic3 Therefore it is consideredthat the S3L inverter has the same output characteristicswith NPC three-level inverter with the same modulationconditions and S3L inverter has a much higher harmoniccontent in output voltage and slightly smaller total distortionrate however its low-order harmonics account for a muchbigger proportion It can be summarized by analyzing Figures12 and 13 that S3L inverter has the same excellent outputwaveforms with NPC3L inverter and its output waveforms ofcurrent are smoothing and approximate sine curve

It can be seen by analyzing Figures 14(a)sim14(f) that theIGBT (1198791015840

1198862

) current surge of S3L inverter is only two-thirds

6 Journal of Electrical and Computer Engineering

Table 2 Commutation processes of S3L inverter

Load current is positive Load current is negativeCommutation Allowed Involved Commutation Allowed Involved1198791015840

1198861

rarr 1198631015840

1198863

11987910158401198863

YES 1198621015840

2

1198631015840

1198861

rarr 1198631015840

1198862

11987910158401198862

YES 1198621015840

2

1198631015840

1198863

11987910158401198863

rarr 1198791015840

1198861

YES 1198621015840

2

1198631015840

1198862

11987910158401198862

rarr 1198631015840

1198861

YES 1198621015840

2

1198631015840

1198863

11987910158401198863

rarr 1198631015840

1198864

YES 1198621015840

1

1198631015840

1198862

11987910158401198862

rarr 1198791015840

1198864

YES 1198621015840

1

1198631015840

1198864

rarr 1198631015840

1198863

11987910158401198863

YES 1198621015840

1

1198791015840

1198864

rarr 1198631015840

1198862

11987910158401198862

YES 1198621015840

1

1198791015840

1198861

rarr 1198631015840

1198864

NO mdash 1198631015840

1198861

rarr 1198791015840

1198864

NO mdash1198631015840

1198864

rarr 1198791015840

1198861

NO mdash 1198791015840

1198864

rarr 1198631015840

1198861

NO mdash

Ud2

Ud2

+

+minus

minus

0

DaT1

DaT2

DaT3

DaT4

L

Load

iL

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(a)

Ud2

Ud2

+

+minus

minus

0

DaT1

DaT2

DaT3

DaT4

L

Load

iL

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(b)

Ud2

Ud2

+

+minus

minus

0

DaT1

DaT2

DaT3

DaT4

L

Load

iL

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(c)

Ud2

Ud2

+

+minus

minus

DaT1

DaT2

DaT3

DaT4

L

Load

iL

0

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(d)

Figure 7 Commutation process of 11987910158401198861

rarr 1198631015840

1198863

11987910158401198863

(a) Before commutation (b) 1199050

le 119905 lt 1199051

period (c) 1199051

le 119905 lt 1199052

period (d) aftercommutation

Table 3 Commutation processes with zero load current

Commutation Allowed Involved1198791015840

1198861

rarr 1198631015840

1198863

11987910158401198863

YES 1198621015840

2

1198631015840

1198863

11987910158401198863

rarr 1198791015840

1198861

YES 1198621015840

2

1198791015840

1198861

rarr 1198631015840

1198864

NO mdash1198631015840

1198861

rarr 1198631015840

1198862

11987910158401198862

YES 1198621015840

1

1198631015840

1198862

11987910158401198862

rarr 1198631015840

1198861

YES 1198621015840

1

1198631015840

1198861

rarr 1198791015840

1198864

YES mdash

of NPC3L inverter (1198791198862) at the switching-on instant the

IGBT (11987910158401198862

) voltage surge of S3L inverter is only half ofNPC3L inverter (119879

1198862) at the switching-off instant overall S3L

Squirrel-cage motor

Double-fed induction

motor

Figure 8 The 21MW dragging platform-motor part

inverters have much lower switching-on and switching-offvoltage and current surges than NPC3L inverters

Journal of Electrical and Computer Engineering 7

NPC3L explosion-proof inverter (1MW) S3L explosion-proof inverter (1MW)

Figure 9 The 21MW dragging platform-inverter part

Time (5msdiv)

CH1

CH2

CH3CH3

(1V

div

)CH

2(1

Vd

iv)

CH1

(1V

div

)

(a) The modulation waveform at119898 = 1154

(b) NPC harmonic content (c) S3L harmonic content

Figure 10 The comparison of phase-voltage harmonic characteristics with119898 = 1154

It is pointed out in the optimized power-loss calculationalgorithm and maternal module concept that it is necessaryto calculate the power losses of other devices except powerdevices on the same bridge to modify the total loss when per-forming the system thermal analysis In this paper themodifi-cation aspects include the power losses of precharge current-limiting resistor balanced resistors absorption capacitanceDC-link capacitors and buffer devices of S3L inverter By theway in any other cases the modification can be calculatedin accordance with the actual conditions These devices areusually fixed in the explosion-proof inverter housing andsome of them are working all the way releasing some power

as constant thermal sources which will elevate the ambienttemperature of the whole cabinet and affect the thermalflow in the cabinet All of this above will finally influencethe power-loss calculation and generate an error betweentheoretical calculation and actual value The conclusion canbe drawn by analyzing Figures 15(a)sim15(d) that the powerloss of IGBT 119879

1198861and its antiparallel diode119863

1198861in NPC three-

level inverters is much bigger than 1198791015840

1198861

and its antiparalleldiode 119863

1015840

1198861

in S3L inverter with different modulations andload impedance angles According to the optimized power-loss calculation algorithm the additional power loss ofeach power device module was calculated and completed

8 Journal of Electrical and Computer Engineering

Time (5msdiv)

CH1

CH2

CH3CH3

(1V

div

)CH

2(1

Vd

iv)

CH1

(1V

div

)(a) The modulation waveform at119898 = 01

(b) NPC harmonic content (c) S3L harmonic content

Figure 11 The comparison of phase-voltage harmonic characteristics with119898 = 01

Time (5msdiv)

CH1 CH2

CH2

(200

Ad

iv)

CH1

(200

Vd

iv)

(a) NPC3L waveforms

Time (5msdiv)

CH1

CH2

CH2

(500

Ad

iv)

CH1

(200

Vd

iv)

(b) S3L waveforms

Figure 12 The comparison of voltage and current in the same IGBT of two inverters

the modification of maternal model power loss in summarythe half-bridge total power loss of S3L inverter maternalmodel is much smaller than NPCrsquos inverter after the mod-ification by 213W on which basis the total power loss ofinverters can be obtained

The calculated power-loss value before modification andaftermodificationwas put into the thermalmodel ofmaternalmodel built by ANSYS ICEPAK respectively as a thermalresource value and the thermal analysis results of twoinverters were presented in Figures 16 and 17 It can be derivedby analyzing the Figures 17(a) and 17(b) that the heat sinktemperature of S3L inverter is about 8∘C lower than that

of NPC three-level inverter running in the same coolingsystems and operating under the same conditions while thesame value in Figures 16(a) and 16(b) before modificationis 3∘C It is easy to find that the substrate temperatureof S3L inverter is 10∘C lower than that of NPC inverterapproximately running in the same cooling systems andoperating under the same conditions by analyzing Figures17(c) and 17(d) But there is a 4∘C decline in Figures 16(c)and 16(d) before modification Overall the analysis showsthat the power devicesrsquo temperature of S3L inverter has a 9∘Cadvantage over NPC inverter under modification and a 35∘Cadvantage without modification

Journal of Electrical and Computer Engineering 9

CH1 AB line voltage (1000Vdiv)

CH2

CH1

CH2 phase A current (500Adiv)

(a) The AB line voltage and phase A current

CH1 AB line voltage (1000Vdiv)

CH1

CH2

CH2 phase A current (500Adiv)

(b) Details of AB line voltage and phase A current

Figure 13 The output voltage and current waveforms of S3L inverter

Time (500nsdiv)

CH1

CH2

CH2

(200

Vd

iv)

CH1

(200

Ad

iv)

(a) IGBT switch-on (NPC)

Time (200nsdiv)

CH1

CH2

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(b) IGBT switch-on (S3L)

Time (200nsdiv)

CH1

CH2

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(c) IGBT switch-off (NPC)

Time (200nsdiv)

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(d) IGBT switch-off (S3L)

Time (200nsdiv)

CH1

(200

Ad

iv)

CH1

(e) NPC diode reverse recovery

Time (200nsdiv)

CH1

CH1

(200

Ad

iv)

(f) S3L diode reverse recovery

Figure 14 The comparative experiments of IGBT and diodersquos characteristics

10 Journal of Electrical and Computer Engineering

500

400

300

200

100

0

Pow

er (W

)

35325

215

105

0

120593 (rad)0

100200

300400

500600

Current (A)

(a) 1198791198861

power loss

400

300

200

100

0

Pow

er (W

)

35325 2

151

050

120593 (rad)0

100 200300

400500

600

Current (A)

(b) 11987910158401198861

power loss

300350

200250

10050

150

0

Powe

r (W

)

35325 2

15105

0

120593 (rad)0 100

200300

400 500600

Current (A)

(c) 1198631198861

power loss

250

300

200

100

50

150

0

Pow

er (W

)

35325

215 1

050

120593 (rad)0 100 200

300400 500

600

Current (A)

(d) 11986310158401198861

power loss

200

150

100

50

0

Pow

er (W

)

35325 2

15105

0 0100

200300

400500

600

Current (A)120593 (rad)

(e) NPC additional power-loss correction

80100120140160180

6040200

Pow

er (W

)

35325 2

15 105

0

120593 (rad)0 100 200

300400

500600

Current (A)

(f) S3L additional power-loss correction

Figure 15The half-bridge power devicesrsquo power-loss calculation of the two inverters based on optimized power-loss algorithm (a)1198791198861

powerloss (b) 1198791015840

1198861

power loss (c) 1198631198861

power loss (d) 11986310158401198861

power loss (e) NPC additional power-loss correction (f) S3L additional power-losscorrection

By observing the experiment results in Figures 18(a)and 18(b) the fact that the heat sink surface temperature ofS3L inverter is about 6∘C lower than that of NPC three-level inverter in average under the same conditions canbe obtained which is in line with the theoretical analysisexpectation by observing the experiment results in Figures18(c) and 18(d) the fact that the IGBT substrate temperatureof S3L inverter is lower than NPC three-level inverter by 11∘Ccan be figured out which is consistent with the theoreticalanalysis results Generally speaking the experimental results

show that the power device temperature of S3L inverter islower than that of NPC three-level inverter by 10∘C approxi-mately which is much closer to the theoretical analysis result9∘C with modification and only has 1∘C error under theseexperiment conditions

In summary the maternal model based on the optimizedpower-loss algorithm has a much higher thermal analysisaccuracy in the improvement process of three-level inverterswhich offers a 1∘C error between theoretical calculationand experiment value in this paper and can be used as

Journal of Electrical and Computer Engineering 11

79828

74736

69643

6455

59458

54365

49272

4418

39087

33994

(a)

7561

70822

66035

61247

5646

51673

46885

42098

37311

32523

(b)

79828

79233

78637

78042

77446

76851

76255

7566

75064

74469

(c)

7561

7502

7443

73841

73251

72662

72072

71483

70893

70303

(d)

Figure 16 The temperature distribution 3D map of maternal thermal models before modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

a tool to support the accurate power-loss calculation andthermal analysis using the S3L inverter based on the softswitching control to improve the NPC three-level invertercan get a good result that S3L inverter has the same excellentoutput characteristic with NPC three-level inverter and hasa great advantage in reducing power loss with a 213Wdecline in each half-bridge and 10∘C decline on power devicetemperatureThe thermal stability of three-level inverters canbe enhanced by this improvement

5 Conclusions

The optimized maternal power-loss thermal models of NPCthree-level inverter and S3L inverter were established based

on the optimized power-loss algorithm and a set of generaloptimized power-loss calculation formulas was derived tomodify the total power loss and figure out the modificationpower-loss values Then these values were considered asthermal sources to analyze the maternal thermal modelsThe three-level inverter can be improved by comparing andanalyzing power-loss modification values and experimentresults Based on this principle and methods the fact thatunder the same conditions the power-lossmodification valueof S3L inverter is smaller than that ofNPC three-level inverterby 213W and has a 9∘C advantage is obtained which is only1∘C smaller than the experiment result Experimental resultsvalidate the proposed theoretical calculation and analysis andprove the effectiveness of the improvement

12 Journal of Electrical and Computer Engineering

61678

58801

55924

53047

5017

47293

44416

41539

38662

35785

(a)

49007

4728

45554

43828

42102

40376

38649

36923

35197

33471

(b)

61678

61374

6107

60766

60462

60159

59855

59551

59247

58943

(c)

49007

48824

48642

4846

48277

48095

47913

47731

47548

47366

(d)

Figure 17 The temperature distribution 3D map of maternal thermal models after modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

Appendix

The parameters of double-fed induction motor used in thisexperiment are as follows

rated power 2100KWrated speed 1000 rpmnumber of pole-pairs 3efficiency at full load 97network voltage 690Vnetwork frequency 50Hzstator current 1662A

rotor current 743Acoupling stator Δ rotor Yrotor open voltage 1710Vinertia 94 kgm2stator maximum short-circuit current 4250Arotor maximum short-circuit current 2900Astator resistance R1 0006013Ωstator leakage reactance X1 0045062Ωrotor resistance (equivalent) R2 0004193Ωrotor leakage reactance (equivalent) X2 02298Ω

Journal of Electrical and Computer Engineering 13

NPC3L

(∘C)

603

56

52

48

44

40

36

32

28

246

(a)

S3L

552

51

48

45

42

39

36

33

30

27

247

(∘C )

(b)

NPC3L

(c)

S3L

(d)

Figure 18The experimental temperature results of IGBTrsquos surface and substrate (a) IGBT surface temperature of NPC3L inverter (b) IGBTsurface temperature of S3L inverter (c) IGBT substrate temperature of NPC3L inverter (d) IGBT substrate temperature of S3L inverter

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors would like to thank 2014 Jiangsu ProvinceNatural Science Foundation (BK20140204) the Research andInnovation Program of Postgraduates in Jiangsu Province(CXZZ13 0930) and the Fundamental Research Funds forthe Central Universities (2012LWB73)

References

[1] S Bernet ldquoRecent developments of high power converters forindustry and traction applicationsrdquo IEEE Transactions on PowerElectronics vol 15 no 6 pp 1102ndash1117 2000

[2] P Mao S-J Xie and Z-G Xu ldquoSwitching transients model andloss analysis of IGBTmodulerdquo Proceedings of the Chinese Societyof Electrical Engineering vol 15 article 007 2010

[3] A D Rajapakse AM Gole and P LWilson ldquoElectromagnetictransients simulation models for accurate representation ofswitching losses and thermal performance in power electronic

systemsrdquo IEEE Transactions on Power Delivery vol 20 no 1 pp319ndash327 2005

[4] J Hu J Li J Zou and J Tan ldquoLosses calculation of IGBTmodule and heat dissipation system design of invertersrdquo Trans-actions of China Electrotechnical Society vol 24 no 3 pp 159ndash163 2009

[5] M H Bierhoff and F W Fuchs ldquoSemiconductor losses involtage source and current source IGBT converters based onanalytical derivationrdquo in Proceedings of the IEEE 35th AnnualPower Electronics Specialists Conference (PESC rsquo04) vol 4 pp2836ndash2842 IEEE June 2004

[6] F Krismer and J W Kolar ldquoAccurate power loss modelderivation of a high-current dual active bridge converter foran automotive applicationrdquo IEEE Transactions on IndustrialElectronics vol 57 no 3 pp 881ndash891 2010

[7] Q Chen Q Wang W Jiang and C Hu ldquoAnalysis of switchinglosses in diode-clamped three-level converterrdquo Transactions ofChina Electrotechnical Society vol 23 no 2 pp 68ndash75 2008

[8] J Wang Q Chen W Jiang and C Hu ldquoAnalysis of conductionlosses in neutral-point-clamped three-level inverterrdquo Transac-tions of China Electrotechnical Society vol 3 p 12 2007

[9] T J Kim DW Kang Y H Lee andD S Hyun ldquoThe analysis ofconduction and switching losses inmulti-level inverter systemrdquoin Proceedings of the IEEE 32nd Annual Power ElectronicsSpecialists Conference (PESC rsquo01) pp 1363ndash1368 June 2001

14 Journal of Electrical and Computer Engineering

[10] S Dieckerhoff S Bernet and D Krug ldquoPower loss-orientedevaluation of high voltage IGBTs and multilevel convertersin transformerless traction applicationsrdquo IEEE Transactions onPower Electronics vol 20 no 6 pp 1328ndash1336 2005

[11] W Jing G Tan and Z Ye ldquoLosses calculation and heat dissipa-tion analysis of high-power three-level convertersrdquoTransactionsof China Electrotechnical Society vol 26 no 2 pp 134ndash140 2011

[12] K Ma Y Yang and F Blaabjerg ldquoTransient modelling ofloss and thermal dynamics in power semiconductor devicesrdquoin Proceedings of the IEEE Energy Conversion Congress andExposition (ECCE rsquo14) pp 5495ndash5501 IEEE 2014

[13] U R Prasanna and A K Rathore ldquoAnalysis design and exper-imental results of a novel soft-switching snubberless current-fed half-bridge front-end converter-based pv inverterrdquo IEEETransactions on Power Electronics vol 28 no 7 pp 3219ndash32302013

[14] X Ruan L Zhou and Y Yan ldquoSoft-switching PWM three-levelconvertersrdquo IEEE Transactions on Power Electronics vol 16 no5 pp 612ndash622 2001

[15] R Li andD Xu ldquoA zero-voltage switching three-phase inverterrdquoIEEE Transactions on Power Electronics vol 29 no 3 pp 1200ndash1210 2014

[16] P Kollensperger R U Lenke S Schroder and R W de Don-cker ldquoDesign of a flexible control platform for soft-switchingmultilevel invertersrdquo IEEE Transactions on Power Electronicsvol 22 no 5 pp 1778ndash1785 2007

[17] G Ortiz H Uemura D Bortis J W Kolar and O ApeldoornldquoModeling of soft-switching losses of IGBTs in high-powerhigh-efficiency dual-active-bridge DCDC convertersrdquo IEEETransactions on Electron Devices vol 60 no 2 pp 587ndash5972013

[18] M W Gekeler ldquoSoft switching three level inverter (S3Linverter)rdquo in Proceedings of the 15th European Conference onPower Electronics and Applications (EPE rsquo13) pp 1ndash10 IEEESeptember 2013

[19] S Munk-Nielsen L N Tutelea and U Jaeger ldquoSimulation withideal switchmodels combinedwithmeasured loss data providesa good estimate of power lossrdquo in Proceedings of the ConferenceRecord of the IEEE Industry Applications Conference vol 5 pp2915ndash2922 Rome Italy October 2000

[20] O S Senturk L Helle S Munk-Nielsen P Rodriguez and RTeodorescu ldquoConverter structure-based power loss and staticthermal modeling of the press-pack IGBT three-level ANPCVSC applied to multi-MW wind turbinesrdquo IEEE Transactionson Industry Applications vol 47 no 6 pp 2505ndash2515 2011

[21] D A B Zambra C Rech F A S Goncalves and J R PinheiroldquoPower losses analysis and cooling system design of threetopologies of multilevel invertersrdquo in Proceedings of the 39thIEEE Annual Power Electronics Specialists Conference (PESCrsquo08) pp 4290ndash4295 IEEE June 2008

[22] W Jing Three-level inverter power loss power devices research[Doctoral dissertation] China University of Mining and Tech-nology Xuzhou China 2011

[23] D Floricau E Floricau and G Gateau ldquoThree-level ActiveNPC converter PWM strategies and loss distributionrdquo inProceedings of the 34th Annual Conference of the IEEE IndustrialElectronics Society (IECON rsquo08) pp 3333ndash3338November 2008

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

6 Journal of Electrical and Computer Engineering

Table 2 Commutation processes of S3L inverter

Load current is positive Load current is negativeCommutation Allowed Involved Commutation Allowed Involved1198791015840

1198861

rarr 1198631015840

1198863

11987910158401198863

YES 1198621015840

2

1198631015840

1198861

rarr 1198631015840

1198862

11987910158401198862

YES 1198621015840

2

1198631015840

1198863

11987910158401198863

rarr 1198791015840

1198861

YES 1198621015840

2

1198631015840

1198862

11987910158401198862

rarr 1198631015840

1198861

YES 1198621015840

2

1198631015840

1198863

11987910158401198863

rarr 1198631015840

1198864

YES 1198621015840

1

1198631015840

1198862

11987910158401198862

rarr 1198791015840

1198864

YES 1198621015840

1

1198631015840

1198864

rarr 1198631015840

1198863

11987910158401198863

YES 1198621015840

1

1198791015840

1198864

rarr 1198631015840

1198862

11987910158401198862

YES 1198621015840

1

1198791015840

1198861

rarr 1198631015840

1198864

NO mdash 1198631015840

1198861

rarr 1198791015840

1198864

NO mdash1198631015840

1198864

rarr 1198791015840

1198861

NO mdash 1198791015840

1198864

rarr 1198631015840

1198861

NO mdash

Ud2

Ud2

+

+minus

minus

0

DaT1

DaT2

DaT3

DaT4

L

Load

iL

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(a)

Ud2

Ud2

+

+minus

minus

0

DaT1

DaT2

DaT3

DaT4

L

Load

iL

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(b)

Ud2

Ud2

+

+minus

minus

0

DaT1

DaT2

DaT3

DaT4

L

Load

iL

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(c)

Ud2

Ud2

+

+minus

minus

DaT1

DaT2

DaT3

DaT4

L

Load

iL

0

C9984001

T998400a1

T998400a2

T998400a3

T998400a4

D998400a2

D998400a3

D998400a4

U998400load

UC9984002

UT9984001198861

(d)

Figure 7 Commutation process of 11987910158401198861

rarr 1198631015840

1198863

11987910158401198863

(a) Before commutation (b) 1199050

le 119905 lt 1199051

period (c) 1199051

le 119905 lt 1199052

period (d) aftercommutation

Table 3 Commutation processes with zero load current

Commutation Allowed Involved1198791015840

1198861

rarr 1198631015840

1198863

11987910158401198863

YES 1198621015840

2

1198631015840

1198863

11987910158401198863

rarr 1198791015840

1198861

YES 1198621015840

2

1198791015840

1198861

rarr 1198631015840

1198864

NO mdash1198631015840

1198861

rarr 1198631015840

1198862

11987910158401198862

YES 1198621015840

1

1198631015840

1198862

11987910158401198862

rarr 1198631015840

1198861

YES 1198621015840

1

1198631015840

1198861

rarr 1198791015840

1198864

YES mdash

of NPC3L inverter (1198791198862) at the switching-on instant the

IGBT (11987910158401198862

) voltage surge of S3L inverter is only half ofNPC3L inverter (119879

1198862) at the switching-off instant overall S3L

Squirrel-cage motor

Double-fed induction

motor

Figure 8 The 21MW dragging platform-motor part

inverters have much lower switching-on and switching-offvoltage and current surges than NPC3L inverters

Journal of Electrical and Computer Engineering 7

NPC3L explosion-proof inverter (1MW) S3L explosion-proof inverter (1MW)

Figure 9 The 21MW dragging platform-inverter part

Time (5msdiv)

CH1

CH2

CH3CH3

(1V

div

)CH

2(1

Vd

iv)

CH1

(1V

div

)

(a) The modulation waveform at119898 = 1154

(b) NPC harmonic content (c) S3L harmonic content

Figure 10 The comparison of phase-voltage harmonic characteristics with119898 = 1154

It is pointed out in the optimized power-loss calculationalgorithm and maternal module concept that it is necessaryto calculate the power losses of other devices except powerdevices on the same bridge to modify the total loss when per-forming the system thermal analysis In this paper themodifi-cation aspects include the power losses of precharge current-limiting resistor balanced resistors absorption capacitanceDC-link capacitors and buffer devices of S3L inverter By theway in any other cases the modification can be calculatedin accordance with the actual conditions These devices areusually fixed in the explosion-proof inverter housing andsome of them are working all the way releasing some power

as constant thermal sources which will elevate the ambienttemperature of the whole cabinet and affect the thermalflow in the cabinet All of this above will finally influencethe power-loss calculation and generate an error betweentheoretical calculation and actual value The conclusion canbe drawn by analyzing Figures 15(a)sim15(d) that the powerloss of IGBT 119879

1198861and its antiparallel diode119863

1198861in NPC three-

level inverters is much bigger than 1198791015840

1198861

and its antiparalleldiode 119863

1015840

1198861

in S3L inverter with different modulations andload impedance angles According to the optimized power-loss calculation algorithm the additional power loss ofeach power device module was calculated and completed

8 Journal of Electrical and Computer Engineering

Time (5msdiv)

CH1

CH2

CH3CH3

(1V

div

)CH

2(1

Vd

iv)

CH1

(1V

div

)(a) The modulation waveform at119898 = 01

(b) NPC harmonic content (c) S3L harmonic content

Figure 11 The comparison of phase-voltage harmonic characteristics with119898 = 01

Time (5msdiv)

CH1 CH2

CH2

(200

Ad

iv)

CH1

(200

Vd

iv)

(a) NPC3L waveforms

Time (5msdiv)

CH1

CH2

CH2

(500

Ad

iv)

CH1

(200

Vd

iv)

(b) S3L waveforms

Figure 12 The comparison of voltage and current in the same IGBT of two inverters

the modification of maternal model power loss in summarythe half-bridge total power loss of S3L inverter maternalmodel is much smaller than NPCrsquos inverter after the mod-ification by 213W on which basis the total power loss ofinverters can be obtained

The calculated power-loss value before modification andaftermodificationwas put into the thermalmodel ofmaternalmodel built by ANSYS ICEPAK respectively as a thermalresource value and the thermal analysis results of twoinverters were presented in Figures 16 and 17 It can be derivedby analyzing the Figures 17(a) and 17(b) that the heat sinktemperature of S3L inverter is about 8∘C lower than that

of NPC three-level inverter running in the same coolingsystems and operating under the same conditions while thesame value in Figures 16(a) and 16(b) before modificationis 3∘C It is easy to find that the substrate temperatureof S3L inverter is 10∘C lower than that of NPC inverterapproximately running in the same cooling systems andoperating under the same conditions by analyzing Figures17(c) and 17(d) But there is a 4∘C decline in Figures 16(c)and 16(d) before modification Overall the analysis showsthat the power devicesrsquo temperature of S3L inverter has a 9∘Cadvantage over NPC inverter under modification and a 35∘Cadvantage without modification

Journal of Electrical and Computer Engineering 9

CH1 AB line voltage (1000Vdiv)

CH2

CH1

CH2 phase A current (500Adiv)

(a) The AB line voltage and phase A current

CH1 AB line voltage (1000Vdiv)

CH1

CH2

CH2 phase A current (500Adiv)

(b) Details of AB line voltage and phase A current

Figure 13 The output voltage and current waveforms of S3L inverter

Time (500nsdiv)

CH1

CH2

CH2

(200

Vd

iv)

CH1

(200

Ad

iv)

(a) IGBT switch-on (NPC)

Time (200nsdiv)

CH1

CH2

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(b) IGBT switch-on (S3L)

Time (200nsdiv)

CH1

CH2

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(c) IGBT switch-off (NPC)

Time (200nsdiv)

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(d) IGBT switch-off (S3L)

Time (200nsdiv)

CH1

(200

Ad

iv)

CH1

(e) NPC diode reverse recovery

Time (200nsdiv)

CH1

CH1

(200

Ad

iv)

(f) S3L diode reverse recovery

Figure 14 The comparative experiments of IGBT and diodersquos characteristics

10 Journal of Electrical and Computer Engineering

500

400

300

200

100

0

Pow

er (W

)

35325

215

105

0

120593 (rad)0

100200

300400

500600

Current (A)

(a) 1198791198861

power loss

400

300

200

100

0

Pow

er (W

)

35325 2

151

050

120593 (rad)0

100 200300

400500

600

Current (A)

(b) 11987910158401198861

power loss

300350

200250

10050

150

0

Powe

r (W

)

35325 2

15105

0

120593 (rad)0 100

200300

400 500600

Current (A)

(c) 1198631198861

power loss

250

300

200

100

50

150

0

Pow

er (W

)

35325

215 1

050

120593 (rad)0 100 200

300400 500

600

Current (A)

(d) 11986310158401198861

power loss

200

150

100

50

0

Pow

er (W

)

35325 2

15105

0 0100

200300

400500

600

Current (A)120593 (rad)

(e) NPC additional power-loss correction

80100120140160180

6040200

Pow

er (W

)

35325 2

15 105

0

120593 (rad)0 100 200

300400

500600

Current (A)

(f) S3L additional power-loss correction

Figure 15The half-bridge power devicesrsquo power-loss calculation of the two inverters based on optimized power-loss algorithm (a)1198791198861

powerloss (b) 1198791015840

1198861

power loss (c) 1198631198861

power loss (d) 11986310158401198861

power loss (e) NPC additional power-loss correction (f) S3L additional power-losscorrection

By observing the experiment results in Figures 18(a)and 18(b) the fact that the heat sink surface temperature ofS3L inverter is about 6∘C lower than that of NPC three-level inverter in average under the same conditions canbe obtained which is in line with the theoretical analysisexpectation by observing the experiment results in Figures18(c) and 18(d) the fact that the IGBT substrate temperatureof S3L inverter is lower than NPC three-level inverter by 11∘Ccan be figured out which is consistent with the theoreticalanalysis results Generally speaking the experimental results

show that the power device temperature of S3L inverter islower than that of NPC three-level inverter by 10∘C approxi-mately which is much closer to the theoretical analysis result9∘C with modification and only has 1∘C error under theseexperiment conditions

In summary the maternal model based on the optimizedpower-loss algorithm has a much higher thermal analysisaccuracy in the improvement process of three-level inverterswhich offers a 1∘C error between theoretical calculationand experiment value in this paper and can be used as

Journal of Electrical and Computer Engineering 11

79828

74736

69643

6455

59458

54365

49272

4418

39087

33994

(a)

7561

70822

66035

61247

5646

51673

46885

42098

37311

32523

(b)

79828

79233

78637

78042

77446

76851

76255

7566

75064

74469

(c)

7561

7502

7443

73841

73251

72662

72072

71483

70893

70303

(d)

Figure 16 The temperature distribution 3D map of maternal thermal models before modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

a tool to support the accurate power-loss calculation andthermal analysis using the S3L inverter based on the softswitching control to improve the NPC three-level invertercan get a good result that S3L inverter has the same excellentoutput characteristic with NPC three-level inverter and hasa great advantage in reducing power loss with a 213Wdecline in each half-bridge and 10∘C decline on power devicetemperatureThe thermal stability of three-level inverters canbe enhanced by this improvement

5 Conclusions

The optimized maternal power-loss thermal models of NPCthree-level inverter and S3L inverter were established based

on the optimized power-loss algorithm and a set of generaloptimized power-loss calculation formulas was derived tomodify the total power loss and figure out the modificationpower-loss values Then these values were considered asthermal sources to analyze the maternal thermal modelsThe three-level inverter can be improved by comparing andanalyzing power-loss modification values and experimentresults Based on this principle and methods the fact thatunder the same conditions the power-lossmodification valueof S3L inverter is smaller than that ofNPC three-level inverterby 213W and has a 9∘C advantage is obtained which is only1∘C smaller than the experiment result Experimental resultsvalidate the proposed theoretical calculation and analysis andprove the effectiveness of the improvement

12 Journal of Electrical and Computer Engineering

61678

58801

55924

53047

5017

47293

44416

41539

38662

35785

(a)

49007

4728

45554

43828

42102

40376

38649

36923

35197

33471

(b)

61678

61374

6107

60766

60462

60159

59855

59551

59247

58943

(c)

49007

48824

48642

4846

48277

48095

47913

47731

47548

47366

(d)

Figure 17 The temperature distribution 3D map of maternal thermal models after modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

Appendix

The parameters of double-fed induction motor used in thisexperiment are as follows

rated power 2100KWrated speed 1000 rpmnumber of pole-pairs 3efficiency at full load 97network voltage 690Vnetwork frequency 50Hzstator current 1662A

rotor current 743Acoupling stator Δ rotor Yrotor open voltage 1710Vinertia 94 kgm2stator maximum short-circuit current 4250Arotor maximum short-circuit current 2900Astator resistance R1 0006013Ωstator leakage reactance X1 0045062Ωrotor resistance (equivalent) R2 0004193Ωrotor leakage reactance (equivalent) X2 02298Ω

Journal of Electrical and Computer Engineering 13

NPC3L

(∘C)

603

56

52

48

44

40

36

32

28

246

(a)

S3L

552

51

48

45

42

39

36

33

30

27

247

(∘C )

(b)

NPC3L

(c)

S3L

(d)

Figure 18The experimental temperature results of IGBTrsquos surface and substrate (a) IGBT surface temperature of NPC3L inverter (b) IGBTsurface temperature of S3L inverter (c) IGBT substrate temperature of NPC3L inverter (d) IGBT substrate temperature of S3L inverter

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors would like to thank 2014 Jiangsu ProvinceNatural Science Foundation (BK20140204) the Research andInnovation Program of Postgraduates in Jiangsu Province(CXZZ13 0930) and the Fundamental Research Funds forthe Central Universities (2012LWB73)

References

[1] S Bernet ldquoRecent developments of high power converters forindustry and traction applicationsrdquo IEEE Transactions on PowerElectronics vol 15 no 6 pp 1102ndash1117 2000

[2] P Mao S-J Xie and Z-G Xu ldquoSwitching transients model andloss analysis of IGBTmodulerdquo Proceedings of the Chinese Societyof Electrical Engineering vol 15 article 007 2010

[3] A D Rajapakse AM Gole and P LWilson ldquoElectromagnetictransients simulation models for accurate representation ofswitching losses and thermal performance in power electronic

systemsrdquo IEEE Transactions on Power Delivery vol 20 no 1 pp319ndash327 2005

[4] J Hu J Li J Zou and J Tan ldquoLosses calculation of IGBTmodule and heat dissipation system design of invertersrdquo Trans-actions of China Electrotechnical Society vol 24 no 3 pp 159ndash163 2009

[5] M H Bierhoff and F W Fuchs ldquoSemiconductor losses involtage source and current source IGBT converters based onanalytical derivationrdquo in Proceedings of the IEEE 35th AnnualPower Electronics Specialists Conference (PESC rsquo04) vol 4 pp2836ndash2842 IEEE June 2004

[6] F Krismer and J W Kolar ldquoAccurate power loss modelderivation of a high-current dual active bridge converter foran automotive applicationrdquo IEEE Transactions on IndustrialElectronics vol 57 no 3 pp 881ndash891 2010

[7] Q Chen Q Wang W Jiang and C Hu ldquoAnalysis of switchinglosses in diode-clamped three-level converterrdquo Transactions ofChina Electrotechnical Society vol 23 no 2 pp 68ndash75 2008

[8] J Wang Q Chen W Jiang and C Hu ldquoAnalysis of conductionlosses in neutral-point-clamped three-level inverterrdquo Transac-tions of China Electrotechnical Society vol 3 p 12 2007

[9] T J Kim DW Kang Y H Lee andD S Hyun ldquoThe analysis ofconduction and switching losses inmulti-level inverter systemrdquoin Proceedings of the IEEE 32nd Annual Power ElectronicsSpecialists Conference (PESC rsquo01) pp 1363ndash1368 June 2001

14 Journal of Electrical and Computer Engineering

[10] S Dieckerhoff S Bernet and D Krug ldquoPower loss-orientedevaluation of high voltage IGBTs and multilevel convertersin transformerless traction applicationsrdquo IEEE Transactions onPower Electronics vol 20 no 6 pp 1328ndash1336 2005

[11] W Jing G Tan and Z Ye ldquoLosses calculation and heat dissipa-tion analysis of high-power three-level convertersrdquoTransactionsof China Electrotechnical Society vol 26 no 2 pp 134ndash140 2011

[12] K Ma Y Yang and F Blaabjerg ldquoTransient modelling ofloss and thermal dynamics in power semiconductor devicesrdquoin Proceedings of the IEEE Energy Conversion Congress andExposition (ECCE rsquo14) pp 5495ndash5501 IEEE 2014

[13] U R Prasanna and A K Rathore ldquoAnalysis design and exper-imental results of a novel soft-switching snubberless current-fed half-bridge front-end converter-based pv inverterrdquo IEEETransactions on Power Electronics vol 28 no 7 pp 3219ndash32302013

[14] X Ruan L Zhou and Y Yan ldquoSoft-switching PWM three-levelconvertersrdquo IEEE Transactions on Power Electronics vol 16 no5 pp 612ndash622 2001

[15] R Li andD Xu ldquoA zero-voltage switching three-phase inverterrdquoIEEE Transactions on Power Electronics vol 29 no 3 pp 1200ndash1210 2014

[16] P Kollensperger R U Lenke S Schroder and R W de Don-cker ldquoDesign of a flexible control platform for soft-switchingmultilevel invertersrdquo IEEE Transactions on Power Electronicsvol 22 no 5 pp 1778ndash1785 2007

[17] G Ortiz H Uemura D Bortis J W Kolar and O ApeldoornldquoModeling of soft-switching losses of IGBTs in high-powerhigh-efficiency dual-active-bridge DCDC convertersrdquo IEEETransactions on Electron Devices vol 60 no 2 pp 587ndash5972013

[18] M W Gekeler ldquoSoft switching three level inverter (S3Linverter)rdquo in Proceedings of the 15th European Conference onPower Electronics and Applications (EPE rsquo13) pp 1ndash10 IEEESeptember 2013

[19] S Munk-Nielsen L N Tutelea and U Jaeger ldquoSimulation withideal switchmodels combinedwithmeasured loss data providesa good estimate of power lossrdquo in Proceedings of the ConferenceRecord of the IEEE Industry Applications Conference vol 5 pp2915ndash2922 Rome Italy October 2000

[20] O S Senturk L Helle S Munk-Nielsen P Rodriguez and RTeodorescu ldquoConverter structure-based power loss and staticthermal modeling of the press-pack IGBT three-level ANPCVSC applied to multi-MW wind turbinesrdquo IEEE Transactionson Industry Applications vol 47 no 6 pp 2505ndash2515 2011

[21] D A B Zambra C Rech F A S Goncalves and J R PinheiroldquoPower losses analysis and cooling system design of threetopologies of multilevel invertersrdquo in Proceedings of the 39thIEEE Annual Power Electronics Specialists Conference (PESCrsquo08) pp 4290ndash4295 IEEE June 2008

[22] W Jing Three-level inverter power loss power devices research[Doctoral dissertation] China University of Mining and Tech-nology Xuzhou China 2011

[23] D Floricau E Floricau and G Gateau ldquoThree-level ActiveNPC converter PWM strategies and loss distributionrdquo inProceedings of the 34th Annual Conference of the IEEE IndustrialElectronics Society (IECON rsquo08) pp 3333ndash3338November 2008

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

Journal of Electrical and Computer Engineering 7

NPC3L explosion-proof inverter (1MW) S3L explosion-proof inverter (1MW)

Figure 9 The 21MW dragging platform-inverter part

Time (5msdiv)

CH1

CH2

CH3CH3

(1V

div

)CH

2(1

Vd

iv)

CH1

(1V

div

)

(a) The modulation waveform at119898 = 1154

(b) NPC harmonic content (c) S3L harmonic content

Figure 10 The comparison of phase-voltage harmonic characteristics with119898 = 1154

It is pointed out in the optimized power-loss calculationalgorithm and maternal module concept that it is necessaryto calculate the power losses of other devices except powerdevices on the same bridge to modify the total loss when per-forming the system thermal analysis In this paper themodifi-cation aspects include the power losses of precharge current-limiting resistor balanced resistors absorption capacitanceDC-link capacitors and buffer devices of S3L inverter By theway in any other cases the modification can be calculatedin accordance with the actual conditions These devices areusually fixed in the explosion-proof inverter housing andsome of them are working all the way releasing some power

as constant thermal sources which will elevate the ambienttemperature of the whole cabinet and affect the thermalflow in the cabinet All of this above will finally influencethe power-loss calculation and generate an error betweentheoretical calculation and actual value The conclusion canbe drawn by analyzing Figures 15(a)sim15(d) that the powerloss of IGBT 119879

1198861and its antiparallel diode119863

1198861in NPC three-

level inverters is much bigger than 1198791015840

1198861

and its antiparalleldiode 119863

1015840

1198861

in S3L inverter with different modulations andload impedance angles According to the optimized power-loss calculation algorithm the additional power loss ofeach power device module was calculated and completed

8 Journal of Electrical and Computer Engineering

Time (5msdiv)

CH1

CH2

CH3CH3

(1V

div

)CH

2(1

Vd

iv)

CH1

(1V

div

)(a) The modulation waveform at119898 = 01

(b) NPC harmonic content (c) S3L harmonic content

Figure 11 The comparison of phase-voltage harmonic characteristics with119898 = 01

Time (5msdiv)

CH1 CH2

CH2

(200

Ad

iv)

CH1

(200

Vd

iv)

(a) NPC3L waveforms

Time (5msdiv)

CH1

CH2

CH2

(500

Ad

iv)

CH1

(200

Vd

iv)

(b) S3L waveforms

Figure 12 The comparison of voltage and current in the same IGBT of two inverters

the modification of maternal model power loss in summarythe half-bridge total power loss of S3L inverter maternalmodel is much smaller than NPCrsquos inverter after the mod-ification by 213W on which basis the total power loss ofinverters can be obtained

The calculated power-loss value before modification andaftermodificationwas put into the thermalmodel ofmaternalmodel built by ANSYS ICEPAK respectively as a thermalresource value and the thermal analysis results of twoinverters were presented in Figures 16 and 17 It can be derivedby analyzing the Figures 17(a) and 17(b) that the heat sinktemperature of S3L inverter is about 8∘C lower than that

of NPC three-level inverter running in the same coolingsystems and operating under the same conditions while thesame value in Figures 16(a) and 16(b) before modificationis 3∘C It is easy to find that the substrate temperatureof S3L inverter is 10∘C lower than that of NPC inverterapproximately running in the same cooling systems andoperating under the same conditions by analyzing Figures17(c) and 17(d) But there is a 4∘C decline in Figures 16(c)and 16(d) before modification Overall the analysis showsthat the power devicesrsquo temperature of S3L inverter has a 9∘Cadvantage over NPC inverter under modification and a 35∘Cadvantage without modification

Journal of Electrical and Computer Engineering 9

CH1 AB line voltage (1000Vdiv)

CH2

CH1

CH2 phase A current (500Adiv)

(a) The AB line voltage and phase A current

CH1 AB line voltage (1000Vdiv)

CH1

CH2

CH2 phase A current (500Adiv)

(b) Details of AB line voltage and phase A current

Figure 13 The output voltage and current waveforms of S3L inverter

Time (500nsdiv)

CH1

CH2

CH2

(200

Vd

iv)

CH1

(200

Ad

iv)

(a) IGBT switch-on (NPC)

Time (200nsdiv)

CH1

CH2

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(b) IGBT switch-on (S3L)

Time (200nsdiv)

CH1

CH2

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(c) IGBT switch-off (NPC)

Time (200nsdiv)

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(d) IGBT switch-off (S3L)

Time (200nsdiv)

CH1

(200

Ad

iv)

CH1

(e) NPC diode reverse recovery

Time (200nsdiv)

CH1

CH1

(200

Ad

iv)

(f) S3L diode reverse recovery

Figure 14 The comparative experiments of IGBT and diodersquos characteristics

10 Journal of Electrical and Computer Engineering

500

400

300

200

100

0

Pow

er (W

)

35325

215

105

0

120593 (rad)0

100200

300400

500600

Current (A)

(a) 1198791198861

power loss

400

300

200

100

0

Pow

er (W

)

35325 2

151

050

120593 (rad)0

100 200300

400500

600

Current (A)

(b) 11987910158401198861

power loss

300350

200250

10050

150

0

Powe

r (W

)

35325 2

15105

0

120593 (rad)0 100

200300

400 500600

Current (A)

(c) 1198631198861

power loss

250

300

200

100

50

150

0

Pow

er (W

)

35325

215 1

050

120593 (rad)0 100 200

300400 500

600

Current (A)

(d) 11986310158401198861

power loss

200

150

100

50

0

Pow

er (W

)

35325 2

15105

0 0100

200300

400500

600

Current (A)120593 (rad)

(e) NPC additional power-loss correction

80100120140160180

6040200

Pow

er (W

)

35325 2

15 105

0

120593 (rad)0 100 200

300400

500600

Current (A)

(f) S3L additional power-loss correction

Figure 15The half-bridge power devicesrsquo power-loss calculation of the two inverters based on optimized power-loss algorithm (a)1198791198861

powerloss (b) 1198791015840

1198861

power loss (c) 1198631198861

power loss (d) 11986310158401198861

power loss (e) NPC additional power-loss correction (f) S3L additional power-losscorrection

By observing the experiment results in Figures 18(a)and 18(b) the fact that the heat sink surface temperature ofS3L inverter is about 6∘C lower than that of NPC three-level inverter in average under the same conditions canbe obtained which is in line with the theoretical analysisexpectation by observing the experiment results in Figures18(c) and 18(d) the fact that the IGBT substrate temperatureof S3L inverter is lower than NPC three-level inverter by 11∘Ccan be figured out which is consistent with the theoreticalanalysis results Generally speaking the experimental results

show that the power device temperature of S3L inverter islower than that of NPC three-level inverter by 10∘C approxi-mately which is much closer to the theoretical analysis result9∘C with modification and only has 1∘C error under theseexperiment conditions

In summary the maternal model based on the optimizedpower-loss algorithm has a much higher thermal analysisaccuracy in the improvement process of three-level inverterswhich offers a 1∘C error between theoretical calculationand experiment value in this paper and can be used as

Journal of Electrical and Computer Engineering 11

79828

74736

69643

6455

59458

54365

49272

4418

39087

33994

(a)

7561

70822

66035

61247

5646

51673

46885

42098

37311

32523

(b)

79828

79233

78637

78042

77446

76851

76255

7566

75064

74469

(c)

7561

7502

7443

73841

73251

72662

72072

71483

70893

70303

(d)

Figure 16 The temperature distribution 3D map of maternal thermal models before modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

a tool to support the accurate power-loss calculation andthermal analysis using the S3L inverter based on the softswitching control to improve the NPC three-level invertercan get a good result that S3L inverter has the same excellentoutput characteristic with NPC three-level inverter and hasa great advantage in reducing power loss with a 213Wdecline in each half-bridge and 10∘C decline on power devicetemperatureThe thermal stability of three-level inverters canbe enhanced by this improvement

5 Conclusions

The optimized maternal power-loss thermal models of NPCthree-level inverter and S3L inverter were established based

on the optimized power-loss algorithm and a set of generaloptimized power-loss calculation formulas was derived tomodify the total power loss and figure out the modificationpower-loss values Then these values were considered asthermal sources to analyze the maternal thermal modelsThe three-level inverter can be improved by comparing andanalyzing power-loss modification values and experimentresults Based on this principle and methods the fact thatunder the same conditions the power-lossmodification valueof S3L inverter is smaller than that ofNPC three-level inverterby 213W and has a 9∘C advantage is obtained which is only1∘C smaller than the experiment result Experimental resultsvalidate the proposed theoretical calculation and analysis andprove the effectiveness of the improvement

12 Journal of Electrical and Computer Engineering

61678

58801

55924

53047

5017

47293

44416

41539

38662

35785

(a)

49007

4728

45554

43828

42102

40376

38649

36923

35197

33471

(b)

61678

61374

6107

60766

60462

60159

59855

59551

59247

58943

(c)

49007

48824

48642

4846

48277

48095

47913

47731

47548

47366

(d)

Figure 17 The temperature distribution 3D map of maternal thermal models after modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

Appendix

The parameters of double-fed induction motor used in thisexperiment are as follows

rated power 2100KWrated speed 1000 rpmnumber of pole-pairs 3efficiency at full load 97network voltage 690Vnetwork frequency 50Hzstator current 1662A

rotor current 743Acoupling stator Δ rotor Yrotor open voltage 1710Vinertia 94 kgm2stator maximum short-circuit current 4250Arotor maximum short-circuit current 2900Astator resistance R1 0006013Ωstator leakage reactance X1 0045062Ωrotor resistance (equivalent) R2 0004193Ωrotor leakage reactance (equivalent) X2 02298Ω

Journal of Electrical and Computer Engineering 13

NPC3L

(∘C)

603

56

52

48

44

40

36

32

28

246

(a)

S3L

552

51

48

45

42

39

36

33

30

27

247

(∘C )

(b)

NPC3L

(c)

S3L

(d)

Figure 18The experimental temperature results of IGBTrsquos surface and substrate (a) IGBT surface temperature of NPC3L inverter (b) IGBTsurface temperature of S3L inverter (c) IGBT substrate temperature of NPC3L inverter (d) IGBT substrate temperature of S3L inverter

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors would like to thank 2014 Jiangsu ProvinceNatural Science Foundation (BK20140204) the Research andInnovation Program of Postgraduates in Jiangsu Province(CXZZ13 0930) and the Fundamental Research Funds forthe Central Universities (2012LWB73)

References

[1] S Bernet ldquoRecent developments of high power converters forindustry and traction applicationsrdquo IEEE Transactions on PowerElectronics vol 15 no 6 pp 1102ndash1117 2000

[2] P Mao S-J Xie and Z-G Xu ldquoSwitching transients model andloss analysis of IGBTmodulerdquo Proceedings of the Chinese Societyof Electrical Engineering vol 15 article 007 2010

[3] A D Rajapakse AM Gole and P LWilson ldquoElectromagnetictransients simulation models for accurate representation ofswitching losses and thermal performance in power electronic

systemsrdquo IEEE Transactions on Power Delivery vol 20 no 1 pp319ndash327 2005

[4] J Hu J Li J Zou and J Tan ldquoLosses calculation of IGBTmodule and heat dissipation system design of invertersrdquo Trans-actions of China Electrotechnical Society vol 24 no 3 pp 159ndash163 2009

[5] M H Bierhoff and F W Fuchs ldquoSemiconductor losses involtage source and current source IGBT converters based onanalytical derivationrdquo in Proceedings of the IEEE 35th AnnualPower Electronics Specialists Conference (PESC rsquo04) vol 4 pp2836ndash2842 IEEE June 2004

[6] F Krismer and J W Kolar ldquoAccurate power loss modelderivation of a high-current dual active bridge converter foran automotive applicationrdquo IEEE Transactions on IndustrialElectronics vol 57 no 3 pp 881ndash891 2010

[7] Q Chen Q Wang W Jiang and C Hu ldquoAnalysis of switchinglosses in diode-clamped three-level converterrdquo Transactions ofChina Electrotechnical Society vol 23 no 2 pp 68ndash75 2008

[8] J Wang Q Chen W Jiang and C Hu ldquoAnalysis of conductionlosses in neutral-point-clamped three-level inverterrdquo Transac-tions of China Electrotechnical Society vol 3 p 12 2007

[9] T J Kim DW Kang Y H Lee andD S Hyun ldquoThe analysis ofconduction and switching losses inmulti-level inverter systemrdquoin Proceedings of the IEEE 32nd Annual Power ElectronicsSpecialists Conference (PESC rsquo01) pp 1363ndash1368 June 2001

14 Journal of Electrical and Computer Engineering

[10] S Dieckerhoff S Bernet and D Krug ldquoPower loss-orientedevaluation of high voltage IGBTs and multilevel convertersin transformerless traction applicationsrdquo IEEE Transactions onPower Electronics vol 20 no 6 pp 1328ndash1336 2005

[11] W Jing G Tan and Z Ye ldquoLosses calculation and heat dissipa-tion analysis of high-power three-level convertersrdquoTransactionsof China Electrotechnical Society vol 26 no 2 pp 134ndash140 2011

[12] K Ma Y Yang and F Blaabjerg ldquoTransient modelling ofloss and thermal dynamics in power semiconductor devicesrdquoin Proceedings of the IEEE Energy Conversion Congress andExposition (ECCE rsquo14) pp 5495ndash5501 IEEE 2014

[13] U R Prasanna and A K Rathore ldquoAnalysis design and exper-imental results of a novel soft-switching snubberless current-fed half-bridge front-end converter-based pv inverterrdquo IEEETransactions on Power Electronics vol 28 no 7 pp 3219ndash32302013

[14] X Ruan L Zhou and Y Yan ldquoSoft-switching PWM three-levelconvertersrdquo IEEE Transactions on Power Electronics vol 16 no5 pp 612ndash622 2001

[15] R Li andD Xu ldquoA zero-voltage switching three-phase inverterrdquoIEEE Transactions on Power Electronics vol 29 no 3 pp 1200ndash1210 2014

[16] P Kollensperger R U Lenke S Schroder and R W de Don-cker ldquoDesign of a flexible control platform for soft-switchingmultilevel invertersrdquo IEEE Transactions on Power Electronicsvol 22 no 5 pp 1778ndash1785 2007

[17] G Ortiz H Uemura D Bortis J W Kolar and O ApeldoornldquoModeling of soft-switching losses of IGBTs in high-powerhigh-efficiency dual-active-bridge DCDC convertersrdquo IEEETransactions on Electron Devices vol 60 no 2 pp 587ndash5972013

[18] M W Gekeler ldquoSoft switching three level inverter (S3Linverter)rdquo in Proceedings of the 15th European Conference onPower Electronics and Applications (EPE rsquo13) pp 1ndash10 IEEESeptember 2013

[19] S Munk-Nielsen L N Tutelea and U Jaeger ldquoSimulation withideal switchmodels combinedwithmeasured loss data providesa good estimate of power lossrdquo in Proceedings of the ConferenceRecord of the IEEE Industry Applications Conference vol 5 pp2915ndash2922 Rome Italy October 2000

[20] O S Senturk L Helle S Munk-Nielsen P Rodriguez and RTeodorescu ldquoConverter structure-based power loss and staticthermal modeling of the press-pack IGBT three-level ANPCVSC applied to multi-MW wind turbinesrdquo IEEE Transactionson Industry Applications vol 47 no 6 pp 2505ndash2515 2011

[21] D A B Zambra C Rech F A S Goncalves and J R PinheiroldquoPower losses analysis and cooling system design of threetopologies of multilevel invertersrdquo in Proceedings of the 39thIEEE Annual Power Electronics Specialists Conference (PESCrsquo08) pp 4290ndash4295 IEEE June 2008

[22] W Jing Three-level inverter power loss power devices research[Doctoral dissertation] China University of Mining and Tech-nology Xuzhou China 2011

[23] D Floricau E Floricau and G Gateau ldquoThree-level ActiveNPC converter PWM strategies and loss distributionrdquo inProceedings of the 34th Annual Conference of the IEEE IndustrialElectronics Society (IECON rsquo08) pp 3333ndash3338November 2008

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

8 Journal of Electrical and Computer Engineering

Time (5msdiv)

CH1

CH2

CH3CH3

(1V

div

)CH

2(1

Vd

iv)

CH1

(1V

div

)(a) The modulation waveform at119898 = 01

(b) NPC harmonic content (c) S3L harmonic content

Figure 11 The comparison of phase-voltage harmonic characteristics with119898 = 01

Time (5msdiv)

CH1 CH2

CH2

(200

Ad

iv)

CH1

(200

Vd

iv)

(a) NPC3L waveforms

Time (5msdiv)

CH1

CH2

CH2

(500

Ad

iv)

CH1

(200

Vd

iv)

(b) S3L waveforms

Figure 12 The comparison of voltage and current in the same IGBT of two inverters

the modification of maternal model power loss in summarythe half-bridge total power loss of S3L inverter maternalmodel is much smaller than NPCrsquos inverter after the mod-ification by 213W on which basis the total power loss ofinverters can be obtained

The calculated power-loss value before modification andaftermodificationwas put into the thermalmodel ofmaternalmodel built by ANSYS ICEPAK respectively as a thermalresource value and the thermal analysis results of twoinverters were presented in Figures 16 and 17 It can be derivedby analyzing the Figures 17(a) and 17(b) that the heat sinktemperature of S3L inverter is about 8∘C lower than that

of NPC three-level inverter running in the same coolingsystems and operating under the same conditions while thesame value in Figures 16(a) and 16(b) before modificationis 3∘C It is easy to find that the substrate temperatureof S3L inverter is 10∘C lower than that of NPC inverterapproximately running in the same cooling systems andoperating under the same conditions by analyzing Figures17(c) and 17(d) But there is a 4∘C decline in Figures 16(c)and 16(d) before modification Overall the analysis showsthat the power devicesrsquo temperature of S3L inverter has a 9∘Cadvantage over NPC inverter under modification and a 35∘Cadvantage without modification

Journal of Electrical and Computer Engineering 9

CH1 AB line voltage (1000Vdiv)

CH2

CH1

CH2 phase A current (500Adiv)

(a) The AB line voltage and phase A current

CH1 AB line voltage (1000Vdiv)

CH1

CH2

CH2 phase A current (500Adiv)

(b) Details of AB line voltage and phase A current

Figure 13 The output voltage and current waveforms of S3L inverter

Time (500nsdiv)

CH1

CH2

CH2

(200

Vd

iv)

CH1

(200

Ad

iv)

(a) IGBT switch-on (NPC)

Time (200nsdiv)

CH1

CH2

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(b) IGBT switch-on (S3L)

Time (200nsdiv)

CH1

CH2

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(c) IGBT switch-off (NPC)

Time (200nsdiv)

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(d) IGBT switch-off (S3L)

Time (200nsdiv)

CH1

(200

Ad

iv)

CH1

(e) NPC diode reverse recovery

Time (200nsdiv)

CH1

CH1

(200

Ad

iv)

(f) S3L diode reverse recovery

Figure 14 The comparative experiments of IGBT and diodersquos characteristics

10 Journal of Electrical and Computer Engineering

500

400

300

200

100

0

Pow

er (W

)

35325

215

105

0

120593 (rad)0

100200

300400

500600

Current (A)

(a) 1198791198861

power loss

400

300

200

100

0

Pow

er (W

)

35325 2

151

050

120593 (rad)0

100 200300

400500

600

Current (A)

(b) 11987910158401198861

power loss

300350

200250

10050

150

0

Powe

r (W

)

35325 2

15105

0

120593 (rad)0 100

200300

400 500600

Current (A)

(c) 1198631198861

power loss

250

300

200

100

50

150

0

Pow

er (W

)

35325

215 1

050

120593 (rad)0 100 200

300400 500

600

Current (A)

(d) 11986310158401198861

power loss

200

150

100

50

0

Pow

er (W

)

35325 2

15105

0 0100

200300

400500

600

Current (A)120593 (rad)

(e) NPC additional power-loss correction

80100120140160180

6040200

Pow

er (W

)

35325 2

15 105

0

120593 (rad)0 100 200

300400

500600

Current (A)

(f) S3L additional power-loss correction

Figure 15The half-bridge power devicesrsquo power-loss calculation of the two inverters based on optimized power-loss algorithm (a)1198791198861

powerloss (b) 1198791015840

1198861

power loss (c) 1198631198861

power loss (d) 11986310158401198861

power loss (e) NPC additional power-loss correction (f) S3L additional power-losscorrection

By observing the experiment results in Figures 18(a)and 18(b) the fact that the heat sink surface temperature ofS3L inverter is about 6∘C lower than that of NPC three-level inverter in average under the same conditions canbe obtained which is in line with the theoretical analysisexpectation by observing the experiment results in Figures18(c) and 18(d) the fact that the IGBT substrate temperatureof S3L inverter is lower than NPC three-level inverter by 11∘Ccan be figured out which is consistent with the theoreticalanalysis results Generally speaking the experimental results

show that the power device temperature of S3L inverter islower than that of NPC three-level inverter by 10∘C approxi-mately which is much closer to the theoretical analysis result9∘C with modification and only has 1∘C error under theseexperiment conditions

In summary the maternal model based on the optimizedpower-loss algorithm has a much higher thermal analysisaccuracy in the improvement process of three-level inverterswhich offers a 1∘C error between theoretical calculationand experiment value in this paper and can be used as

Journal of Electrical and Computer Engineering 11

79828

74736

69643

6455

59458

54365

49272

4418

39087

33994

(a)

7561

70822

66035

61247

5646

51673

46885

42098

37311

32523

(b)

79828

79233

78637

78042

77446

76851

76255

7566

75064

74469

(c)

7561

7502

7443

73841

73251

72662

72072

71483

70893

70303

(d)

Figure 16 The temperature distribution 3D map of maternal thermal models before modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

a tool to support the accurate power-loss calculation andthermal analysis using the S3L inverter based on the softswitching control to improve the NPC three-level invertercan get a good result that S3L inverter has the same excellentoutput characteristic with NPC three-level inverter and hasa great advantage in reducing power loss with a 213Wdecline in each half-bridge and 10∘C decline on power devicetemperatureThe thermal stability of three-level inverters canbe enhanced by this improvement

5 Conclusions

The optimized maternal power-loss thermal models of NPCthree-level inverter and S3L inverter were established based

on the optimized power-loss algorithm and a set of generaloptimized power-loss calculation formulas was derived tomodify the total power loss and figure out the modificationpower-loss values Then these values were considered asthermal sources to analyze the maternal thermal modelsThe three-level inverter can be improved by comparing andanalyzing power-loss modification values and experimentresults Based on this principle and methods the fact thatunder the same conditions the power-lossmodification valueof S3L inverter is smaller than that ofNPC three-level inverterby 213W and has a 9∘C advantage is obtained which is only1∘C smaller than the experiment result Experimental resultsvalidate the proposed theoretical calculation and analysis andprove the effectiveness of the improvement

12 Journal of Electrical and Computer Engineering

61678

58801

55924

53047

5017

47293

44416

41539

38662

35785

(a)

49007

4728

45554

43828

42102

40376

38649

36923

35197

33471

(b)

61678

61374

6107

60766

60462

60159

59855

59551

59247

58943

(c)

49007

48824

48642

4846

48277

48095

47913

47731

47548

47366

(d)

Figure 17 The temperature distribution 3D map of maternal thermal models after modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

Appendix

The parameters of double-fed induction motor used in thisexperiment are as follows

rated power 2100KWrated speed 1000 rpmnumber of pole-pairs 3efficiency at full load 97network voltage 690Vnetwork frequency 50Hzstator current 1662A

rotor current 743Acoupling stator Δ rotor Yrotor open voltage 1710Vinertia 94 kgm2stator maximum short-circuit current 4250Arotor maximum short-circuit current 2900Astator resistance R1 0006013Ωstator leakage reactance X1 0045062Ωrotor resistance (equivalent) R2 0004193Ωrotor leakage reactance (equivalent) X2 02298Ω

Journal of Electrical and Computer Engineering 13

NPC3L

(∘C)

603

56

52

48

44

40

36

32

28

246

(a)

S3L

552

51

48

45

42

39

36

33

30

27

247

(∘C )

(b)

NPC3L

(c)

S3L

(d)

Figure 18The experimental temperature results of IGBTrsquos surface and substrate (a) IGBT surface temperature of NPC3L inverter (b) IGBTsurface temperature of S3L inverter (c) IGBT substrate temperature of NPC3L inverter (d) IGBT substrate temperature of S3L inverter

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors would like to thank 2014 Jiangsu ProvinceNatural Science Foundation (BK20140204) the Research andInnovation Program of Postgraduates in Jiangsu Province(CXZZ13 0930) and the Fundamental Research Funds forthe Central Universities (2012LWB73)

References

[1] S Bernet ldquoRecent developments of high power converters forindustry and traction applicationsrdquo IEEE Transactions on PowerElectronics vol 15 no 6 pp 1102ndash1117 2000

[2] P Mao S-J Xie and Z-G Xu ldquoSwitching transients model andloss analysis of IGBTmodulerdquo Proceedings of the Chinese Societyof Electrical Engineering vol 15 article 007 2010

[3] A D Rajapakse AM Gole and P LWilson ldquoElectromagnetictransients simulation models for accurate representation ofswitching losses and thermal performance in power electronic

systemsrdquo IEEE Transactions on Power Delivery vol 20 no 1 pp319ndash327 2005

[4] J Hu J Li J Zou and J Tan ldquoLosses calculation of IGBTmodule and heat dissipation system design of invertersrdquo Trans-actions of China Electrotechnical Society vol 24 no 3 pp 159ndash163 2009

[5] M H Bierhoff and F W Fuchs ldquoSemiconductor losses involtage source and current source IGBT converters based onanalytical derivationrdquo in Proceedings of the IEEE 35th AnnualPower Electronics Specialists Conference (PESC rsquo04) vol 4 pp2836ndash2842 IEEE June 2004

[6] F Krismer and J W Kolar ldquoAccurate power loss modelderivation of a high-current dual active bridge converter foran automotive applicationrdquo IEEE Transactions on IndustrialElectronics vol 57 no 3 pp 881ndash891 2010

[7] Q Chen Q Wang W Jiang and C Hu ldquoAnalysis of switchinglosses in diode-clamped three-level converterrdquo Transactions ofChina Electrotechnical Society vol 23 no 2 pp 68ndash75 2008

[8] J Wang Q Chen W Jiang and C Hu ldquoAnalysis of conductionlosses in neutral-point-clamped three-level inverterrdquo Transac-tions of China Electrotechnical Society vol 3 p 12 2007

[9] T J Kim DW Kang Y H Lee andD S Hyun ldquoThe analysis ofconduction and switching losses inmulti-level inverter systemrdquoin Proceedings of the IEEE 32nd Annual Power ElectronicsSpecialists Conference (PESC rsquo01) pp 1363ndash1368 June 2001

14 Journal of Electrical and Computer Engineering

[10] S Dieckerhoff S Bernet and D Krug ldquoPower loss-orientedevaluation of high voltage IGBTs and multilevel convertersin transformerless traction applicationsrdquo IEEE Transactions onPower Electronics vol 20 no 6 pp 1328ndash1336 2005

[11] W Jing G Tan and Z Ye ldquoLosses calculation and heat dissipa-tion analysis of high-power three-level convertersrdquoTransactionsof China Electrotechnical Society vol 26 no 2 pp 134ndash140 2011

[12] K Ma Y Yang and F Blaabjerg ldquoTransient modelling ofloss and thermal dynamics in power semiconductor devicesrdquoin Proceedings of the IEEE Energy Conversion Congress andExposition (ECCE rsquo14) pp 5495ndash5501 IEEE 2014

[13] U R Prasanna and A K Rathore ldquoAnalysis design and exper-imental results of a novel soft-switching snubberless current-fed half-bridge front-end converter-based pv inverterrdquo IEEETransactions on Power Electronics vol 28 no 7 pp 3219ndash32302013

[14] X Ruan L Zhou and Y Yan ldquoSoft-switching PWM three-levelconvertersrdquo IEEE Transactions on Power Electronics vol 16 no5 pp 612ndash622 2001

[15] R Li andD Xu ldquoA zero-voltage switching three-phase inverterrdquoIEEE Transactions on Power Electronics vol 29 no 3 pp 1200ndash1210 2014

[16] P Kollensperger R U Lenke S Schroder and R W de Don-cker ldquoDesign of a flexible control platform for soft-switchingmultilevel invertersrdquo IEEE Transactions on Power Electronicsvol 22 no 5 pp 1778ndash1785 2007

[17] G Ortiz H Uemura D Bortis J W Kolar and O ApeldoornldquoModeling of soft-switching losses of IGBTs in high-powerhigh-efficiency dual-active-bridge DCDC convertersrdquo IEEETransactions on Electron Devices vol 60 no 2 pp 587ndash5972013

[18] M W Gekeler ldquoSoft switching three level inverter (S3Linverter)rdquo in Proceedings of the 15th European Conference onPower Electronics and Applications (EPE rsquo13) pp 1ndash10 IEEESeptember 2013

[19] S Munk-Nielsen L N Tutelea and U Jaeger ldquoSimulation withideal switchmodels combinedwithmeasured loss data providesa good estimate of power lossrdquo in Proceedings of the ConferenceRecord of the IEEE Industry Applications Conference vol 5 pp2915ndash2922 Rome Italy October 2000

[20] O S Senturk L Helle S Munk-Nielsen P Rodriguez and RTeodorescu ldquoConverter structure-based power loss and staticthermal modeling of the press-pack IGBT three-level ANPCVSC applied to multi-MW wind turbinesrdquo IEEE Transactionson Industry Applications vol 47 no 6 pp 2505ndash2515 2011

[21] D A B Zambra C Rech F A S Goncalves and J R PinheiroldquoPower losses analysis and cooling system design of threetopologies of multilevel invertersrdquo in Proceedings of the 39thIEEE Annual Power Electronics Specialists Conference (PESCrsquo08) pp 4290ndash4295 IEEE June 2008

[22] W Jing Three-level inverter power loss power devices research[Doctoral dissertation] China University of Mining and Tech-nology Xuzhou China 2011

[23] D Floricau E Floricau and G Gateau ldquoThree-level ActiveNPC converter PWM strategies and loss distributionrdquo inProceedings of the 34th Annual Conference of the IEEE IndustrialElectronics Society (IECON rsquo08) pp 3333ndash3338November 2008

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

Journal of Electrical and Computer Engineering 9

CH1 AB line voltage (1000Vdiv)

CH2

CH1

CH2 phase A current (500Adiv)

(a) The AB line voltage and phase A current

CH1 AB line voltage (1000Vdiv)

CH1

CH2

CH2 phase A current (500Adiv)

(b) Details of AB line voltage and phase A current

Figure 13 The output voltage and current waveforms of S3L inverter

Time (500nsdiv)

CH1

CH2

CH2

(200

Vd

iv)

CH1

(200

Ad

iv)

(a) IGBT switch-on (NPC)

Time (200nsdiv)

CH1

CH2

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(b) IGBT switch-on (S3L)

Time (200nsdiv)

CH1

CH2

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(c) IGBT switch-off (NPC)

Time (200nsdiv)

CH1

(200

Vd

iv)

CH2

(200

Ad

iv)

(d) IGBT switch-off (S3L)

Time (200nsdiv)

CH1

(200

Ad

iv)

CH1

(e) NPC diode reverse recovery

Time (200nsdiv)

CH1

CH1

(200

Ad

iv)

(f) S3L diode reverse recovery

Figure 14 The comparative experiments of IGBT and diodersquos characteristics

10 Journal of Electrical and Computer Engineering

500

400

300

200

100

0

Pow

er (W

)

35325

215

105

0

120593 (rad)0

100200

300400

500600

Current (A)

(a) 1198791198861

power loss

400

300

200

100

0

Pow

er (W

)

35325 2

151

050

120593 (rad)0

100 200300

400500

600

Current (A)

(b) 11987910158401198861

power loss

300350

200250

10050

150

0

Powe

r (W

)

35325 2

15105

0

120593 (rad)0 100

200300

400 500600

Current (A)

(c) 1198631198861

power loss

250

300

200

100

50

150

0

Pow

er (W

)

35325

215 1

050

120593 (rad)0 100 200

300400 500

600

Current (A)

(d) 11986310158401198861

power loss

200

150

100

50

0

Pow

er (W

)

35325 2

15105

0 0100

200300

400500

600

Current (A)120593 (rad)

(e) NPC additional power-loss correction

80100120140160180

6040200

Pow

er (W

)

35325 2

15 105

0

120593 (rad)0 100 200

300400

500600

Current (A)

(f) S3L additional power-loss correction

Figure 15The half-bridge power devicesrsquo power-loss calculation of the two inverters based on optimized power-loss algorithm (a)1198791198861

powerloss (b) 1198791015840

1198861

power loss (c) 1198631198861

power loss (d) 11986310158401198861

power loss (e) NPC additional power-loss correction (f) S3L additional power-losscorrection

By observing the experiment results in Figures 18(a)and 18(b) the fact that the heat sink surface temperature ofS3L inverter is about 6∘C lower than that of NPC three-level inverter in average under the same conditions canbe obtained which is in line with the theoretical analysisexpectation by observing the experiment results in Figures18(c) and 18(d) the fact that the IGBT substrate temperatureof S3L inverter is lower than NPC three-level inverter by 11∘Ccan be figured out which is consistent with the theoreticalanalysis results Generally speaking the experimental results

show that the power device temperature of S3L inverter islower than that of NPC three-level inverter by 10∘C approxi-mately which is much closer to the theoretical analysis result9∘C with modification and only has 1∘C error under theseexperiment conditions

In summary the maternal model based on the optimizedpower-loss algorithm has a much higher thermal analysisaccuracy in the improvement process of three-level inverterswhich offers a 1∘C error between theoretical calculationand experiment value in this paper and can be used as

Journal of Electrical and Computer Engineering 11

79828

74736

69643

6455

59458

54365

49272

4418

39087

33994

(a)

7561

70822

66035

61247

5646

51673

46885

42098

37311

32523

(b)

79828

79233

78637

78042

77446

76851

76255

7566

75064

74469

(c)

7561

7502

7443

73841

73251

72662

72072

71483

70893

70303

(d)

Figure 16 The temperature distribution 3D map of maternal thermal models before modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

a tool to support the accurate power-loss calculation andthermal analysis using the S3L inverter based on the softswitching control to improve the NPC three-level invertercan get a good result that S3L inverter has the same excellentoutput characteristic with NPC three-level inverter and hasa great advantage in reducing power loss with a 213Wdecline in each half-bridge and 10∘C decline on power devicetemperatureThe thermal stability of three-level inverters canbe enhanced by this improvement

5 Conclusions

The optimized maternal power-loss thermal models of NPCthree-level inverter and S3L inverter were established based

on the optimized power-loss algorithm and a set of generaloptimized power-loss calculation formulas was derived tomodify the total power loss and figure out the modificationpower-loss values Then these values were considered asthermal sources to analyze the maternal thermal modelsThe three-level inverter can be improved by comparing andanalyzing power-loss modification values and experimentresults Based on this principle and methods the fact thatunder the same conditions the power-lossmodification valueof S3L inverter is smaller than that ofNPC three-level inverterby 213W and has a 9∘C advantage is obtained which is only1∘C smaller than the experiment result Experimental resultsvalidate the proposed theoretical calculation and analysis andprove the effectiveness of the improvement

12 Journal of Electrical and Computer Engineering

61678

58801

55924

53047

5017

47293

44416

41539

38662

35785

(a)

49007

4728

45554

43828

42102

40376

38649

36923

35197

33471

(b)

61678

61374

6107

60766

60462

60159

59855

59551

59247

58943

(c)

49007

48824

48642

4846

48277

48095

47913

47731

47548

47366

(d)

Figure 17 The temperature distribution 3D map of maternal thermal models after modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

Appendix

The parameters of double-fed induction motor used in thisexperiment are as follows

rated power 2100KWrated speed 1000 rpmnumber of pole-pairs 3efficiency at full load 97network voltage 690Vnetwork frequency 50Hzstator current 1662A

rotor current 743Acoupling stator Δ rotor Yrotor open voltage 1710Vinertia 94 kgm2stator maximum short-circuit current 4250Arotor maximum short-circuit current 2900Astator resistance R1 0006013Ωstator leakage reactance X1 0045062Ωrotor resistance (equivalent) R2 0004193Ωrotor leakage reactance (equivalent) X2 02298Ω

Journal of Electrical and Computer Engineering 13

NPC3L

(∘C)

603

56

52

48

44

40

36

32

28

246

(a)

S3L

552

51

48

45

42

39

36

33

30

27

247

(∘C )

(b)

NPC3L

(c)

S3L

(d)

Figure 18The experimental temperature results of IGBTrsquos surface and substrate (a) IGBT surface temperature of NPC3L inverter (b) IGBTsurface temperature of S3L inverter (c) IGBT substrate temperature of NPC3L inverter (d) IGBT substrate temperature of S3L inverter

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors would like to thank 2014 Jiangsu ProvinceNatural Science Foundation (BK20140204) the Research andInnovation Program of Postgraduates in Jiangsu Province(CXZZ13 0930) and the Fundamental Research Funds forthe Central Universities (2012LWB73)

References

[1] S Bernet ldquoRecent developments of high power converters forindustry and traction applicationsrdquo IEEE Transactions on PowerElectronics vol 15 no 6 pp 1102ndash1117 2000

[2] P Mao S-J Xie and Z-G Xu ldquoSwitching transients model andloss analysis of IGBTmodulerdquo Proceedings of the Chinese Societyof Electrical Engineering vol 15 article 007 2010

[3] A D Rajapakse AM Gole and P LWilson ldquoElectromagnetictransients simulation models for accurate representation ofswitching losses and thermal performance in power electronic

systemsrdquo IEEE Transactions on Power Delivery vol 20 no 1 pp319ndash327 2005

[4] J Hu J Li J Zou and J Tan ldquoLosses calculation of IGBTmodule and heat dissipation system design of invertersrdquo Trans-actions of China Electrotechnical Society vol 24 no 3 pp 159ndash163 2009

[5] M H Bierhoff and F W Fuchs ldquoSemiconductor losses involtage source and current source IGBT converters based onanalytical derivationrdquo in Proceedings of the IEEE 35th AnnualPower Electronics Specialists Conference (PESC rsquo04) vol 4 pp2836ndash2842 IEEE June 2004

[6] F Krismer and J W Kolar ldquoAccurate power loss modelderivation of a high-current dual active bridge converter foran automotive applicationrdquo IEEE Transactions on IndustrialElectronics vol 57 no 3 pp 881ndash891 2010

[7] Q Chen Q Wang W Jiang and C Hu ldquoAnalysis of switchinglosses in diode-clamped three-level converterrdquo Transactions ofChina Electrotechnical Society vol 23 no 2 pp 68ndash75 2008

[8] J Wang Q Chen W Jiang and C Hu ldquoAnalysis of conductionlosses in neutral-point-clamped three-level inverterrdquo Transac-tions of China Electrotechnical Society vol 3 p 12 2007

[9] T J Kim DW Kang Y H Lee andD S Hyun ldquoThe analysis ofconduction and switching losses inmulti-level inverter systemrdquoin Proceedings of the IEEE 32nd Annual Power ElectronicsSpecialists Conference (PESC rsquo01) pp 1363ndash1368 June 2001

14 Journal of Electrical and Computer Engineering

[10] S Dieckerhoff S Bernet and D Krug ldquoPower loss-orientedevaluation of high voltage IGBTs and multilevel convertersin transformerless traction applicationsrdquo IEEE Transactions onPower Electronics vol 20 no 6 pp 1328ndash1336 2005

[11] W Jing G Tan and Z Ye ldquoLosses calculation and heat dissipa-tion analysis of high-power three-level convertersrdquoTransactionsof China Electrotechnical Society vol 26 no 2 pp 134ndash140 2011

[12] K Ma Y Yang and F Blaabjerg ldquoTransient modelling ofloss and thermal dynamics in power semiconductor devicesrdquoin Proceedings of the IEEE Energy Conversion Congress andExposition (ECCE rsquo14) pp 5495ndash5501 IEEE 2014

[13] U R Prasanna and A K Rathore ldquoAnalysis design and exper-imental results of a novel soft-switching snubberless current-fed half-bridge front-end converter-based pv inverterrdquo IEEETransactions on Power Electronics vol 28 no 7 pp 3219ndash32302013

[14] X Ruan L Zhou and Y Yan ldquoSoft-switching PWM three-levelconvertersrdquo IEEE Transactions on Power Electronics vol 16 no5 pp 612ndash622 2001

[15] R Li andD Xu ldquoA zero-voltage switching three-phase inverterrdquoIEEE Transactions on Power Electronics vol 29 no 3 pp 1200ndash1210 2014

[16] P Kollensperger R U Lenke S Schroder and R W de Don-cker ldquoDesign of a flexible control platform for soft-switchingmultilevel invertersrdquo IEEE Transactions on Power Electronicsvol 22 no 5 pp 1778ndash1785 2007

[17] G Ortiz H Uemura D Bortis J W Kolar and O ApeldoornldquoModeling of soft-switching losses of IGBTs in high-powerhigh-efficiency dual-active-bridge DCDC convertersrdquo IEEETransactions on Electron Devices vol 60 no 2 pp 587ndash5972013

[18] M W Gekeler ldquoSoft switching three level inverter (S3Linverter)rdquo in Proceedings of the 15th European Conference onPower Electronics and Applications (EPE rsquo13) pp 1ndash10 IEEESeptember 2013

[19] S Munk-Nielsen L N Tutelea and U Jaeger ldquoSimulation withideal switchmodels combinedwithmeasured loss data providesa good estimate of power lossrdquo in Proceedings of the ConferenceRecord of the IEEE Industry Applications Conference vol 5 pp2915ndash2922 Rome Italy October 2000

[20] O S Senturk L Helle S Munk-Nielsen P Rodriguez and RTeodorescu ldquoConverter structure-based power loss and staticthermal modeling of the press-pack IGBT three-level ANPCVSC applied to multi-MW wind turbinesrdquo IEEE Transactionson Industry Applications vol 47 no 6 pp 2505ndash2515 2011

[21] D A B Zambra C Rech F A S Goncalves and J R PinheiroldquoPower losses analysis and cooling system design of threetopologies of multilevel invertersrdquo in Proceedings of the 39thIEEE Annual Power Electronics Specialists Conference (PESCrsquo08) pp 4290ndash4295 IEEE June 2008

[22] W Jing Three-level inverter power loss power devices research[Doctoral dissertation] China University of Mining and Tech-nology Xuzhou China 2011

[23] D Floricau E Floricau and G Gateau ldquoThree-level ActiveNPC converter PWM strategies and loss distributionrdquo inProceedings of the 34th Annual Conference of the IEEE IndustrialElectronics Society (IECON rsquo08) pp 3333ndash3338November 2008

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

10 Journal of Electrical and Computer Engineering

500

400

300

200

100

0

Pow

er (W

)

35325

215

105

0

120593 (rad)0

100200

300400

500600

Current (A)

(a) 1198791198861

power loss

400

300

200

100

0

Pow

er (W

)

35325 2

151

050

120593 (rad)0

100 200300

400500

600

Current (A)

(b) 11987910158401198861

power loss

300350

200250

10050

150

0

Powe

r (W

)

35325 2

15105

0

120593 (rad)0 100

200300

400 500600

Current (A)

(c) 1198631198861

power loss

250

300

200

100

50

150

0

Pow

er (W

)

35325

215 1

050

120593 (rad)0 100 200

300400 500

600

Current (A)

(d) 11986310158401198861

power loss

200

150

100

50

0

Pow

er (W

)

35325 2

15105

0 0100

200300

400500

600

Current (A)120593 (rad)

(e) NPC additional power-loss correction

80100120140160180

6040200

Pow

er (W

)

35325 2

15 105

0

120593 (rad)0 100 200

300400

500600

Current (A)

(f) S3L additional power-loss correction

Figure 15The half-bridge power devicesrsquo power-loss calculation of the two inverters based on optimized power-loss algorithm (a)1198791198861

powerloss (b) 1198791015840

1198861

power loss (c) 1198631198861

power loss (d) 11986310158401198861

power loss (e) NPC additional power-loss correction (f) S3L additional power-losscorrection

By observing the experiment results in Figures 18(a)and 18(b) the fact that the heat sink surface temperature ofS3L inverter is about 6∘C lower than that of NPC three-level inverter in average under the same conditions canbe obtained which is in line with the theoretical analysisexpectation by observing the experiment results in Figures18(c) and 18(d) the fact that the IGBT substrate temperatureof S3L inverter is lower than NPC three-level inverter by 11∘Ccan be figured out which is consistent with the theoreticalanalysis results Generally speaking the experimental results

show that the power device temperature of S3L inverter islower than that of NPC three-level inverter by 10∘C approxi-mately which is much closer to the theoretical analysis result9∘C with modification and only has 1∘C error under theseexperiment conditions

In summary the maternal model based on the optimizedpower-loss algorithm has a much higher thermal analysisaccuracy in the improvement process of three-level inverterswhich offers a 1∘C error between theoretical calculationand experiment value in this paper and can be used as

Journal of Electrical and Computer Engineering 11

79828

74736

69643

6455

59458

54365

49272

4418

39087

33994

(a)

7561

70822

66035

61247

5646

51673

46885

42098

37311

32523

(b)

79828

79233

78637

78042

77446

76851

76255

7566

75064

74469

(c)

7561

7502

7443

73841

73251

72662

72072

71483

70893

70303

(d)

Figure 16 The temperature distribution 3D map of maternal thermal models before modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

a tool to support the accurate power-loss calculation andthermal analysis using the S3L inverter based on the softswitching control to improve the NPC three-level invertercan get a good result that S3L inverter has the same excellentoutput characteristic with NPC three-level inverter and hasa great advantage in reducing power loss with a 213Wdecline in each half-bridge and 10∘C decline on power devicetemperatureThe thermal stability of three-level inverters canbe enhanced by this improvement

5 Conclusions

The optimized maternal power-loss thermal models of NPCthree-level inverter and S3L inverter were established based

on the optimized power-loss algorithm and a set of generaloptimized power-loss calculation formulas was derived tomodify the total power loss and figure out the modificationpower-loss values Then these values were considered asthermal sources to analyze the maternal thermal modelsThe three-level inverter can be improved by comparing andanalyzing power-loss modification values and experimentresults Based on this principle and methods the fact thatunder the same conditions the power-lossmodification valueof S3L inverter is smaller than that ofNPC three-level inverterby 213W and has a 9∘C advantage is obtained which is only1∘C smaller than the experiment result Experimental resultsvalidate the proposed theoretical calculation and analysis andprove the effectiveness of the improvement

12 Journal of Electrical and Computer Engineering

61678

58801

55924

53047

5017

47293

44416

41539

38662

35785

(a)

49007

4728

45554

43828

42102

40376

38649

36923

35197

33471

(b)

61678

61374

6107

60766

60462

60159

59855

59551

59247

58943

(c)

49007

48824

48642

4846

48277

48095

47913

47731

47548

47366

(d)

Figure 17 The temperature distribution 3D map of maternal thermal models after modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

Appendix

The parameters of double-fed induction motor used in thisexperiment are as follows

rated power 2100KWrated speed 1000 rpmnumber of pole-pairs 3efficiency at full load 97network voltage 690Vnetwork frequency 50Hzstator current 1662A

rotor current 743Acoupling stator Δ rotor Yrotor open voltage 1710Vinertia 94 kgm2stator maximum short-circuit current 4250Arotor maximum short-circuit current 2900Astator resistance R1 0006013Ωstator leakage reactance X1 0045062Ωrotor resistance (equivalent) R2 0004193Ωrotor leakage reactance (equivalent) X2 02298Ω

Journal of Electrical and Computer Engineering 13

NPC3L

(∘C)

603

56

52

48

44

40

36

32

28

246

(a)

S3L

552

51

48

45

42

39

36

33

30

27

247

(∘C )

(b)

NPC3L

(c)

S3L

(d)

Figure 18The experimental temperature results of IGBTrsquos surface and substrate (a) IGBT surface temperature of NPC3L inverter (b) IGBTsurface temperature of S3L inverter (c) IGBT substrate temperature of NPC3L inverter (d) IGBT substrate temperature of S3L inverter

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors would like to thank 2014 Jiangsu ProvinceNatural Science Foundation (BK20140204) the Research andInnovation Program of Postgraduates in Jiangsu Province(CXZZ13 0930) and the Fundamental Research Funds forthe Central Universities (2012LWB73)

References

[1] S Bernet ldquoRecent developments of high power converters forindustry and traction applicationsrdquo IEEE Transactions on PowerElectronics vol 15 no 6 pp 1102ndash1117 2000

[2] P Mao S-J Xie and Z-G Xu ldquoSwitching transients model andloss analysis of IGBTmodulerdquo Proceedings of the Chinese Societyof Electrical Engineering vol 15 article 007 2010

[3] A D Rajapakse AM Gole and P LWilson ldquoElectromagnetictransients simulation models for accurate representation ofswitching losses and thermal performance in power electronic

systemsrdquo IEEE Transactions on Power Delivery vol 20 no 1 pp319ndash327 2005

[4] J Hu J Li J Zou and J Tan ldquoLosses calculation of IGBTmodule and heat dissipation system design of invertersrdquo Trans-actions of China Electrotechnical Society vol 24 no 3 pp 159ndash163 2009

[5] M H Bierhoff and F W Fuchs ldquoSemiconductor losses involtage source and current source IGBT converters based onanalytical derivationrdquo in Proceedings of the IEEE 35th AnnualPower Electronics Specialists Conference (PESC rsquo04) vol 4 pp2836ndash2842 IEEE June 2004

[6] F Krismer and J W Kolar ldquoAccurate power loss modelderivation of a high-current dual active bridge converter foran automotive applicationrdquo IEEE Transactions on IndustrialElectronics vol 57 no 3 pp 881ndash891 2010

[7] Q Chen Q Wang W Jiang and C Hu ldquoAnalysis of switchinglosses in diode-clamped three-level converterrdquo Transactions ofChina Electrotechnical Society vol 23 no 2 pp 68ndash75 2008

[8] J Wang Q Chen W Jiang and C Hu ldquoAnalysis of conductionlosses in neutral-point-clamped three-level inverterrdquo Transac-tions of China Electrotechnical Society vol 3 p 12 2007

[9] T J Kim DW Kang Y H Lee andD S Hyun ldquoThe analysis ofconduction and switching losses inmulti-level inverter systemrdquoin Proceedings of the IEEE 32nd Annual Power ElectronicsSpecialists Conference (PESC rsquo01) pp 1363ndash1368 June 2001

14 Journal of Electrical and Computer Engineering

[10] S Dieckerhoff S Bernet and D Krug ldquoPower loss-orientedevaluation of high voltage IGBTs and multilevel convertersin transformerless traction applicationsrdquo IEEE Transactions onPower Electronics vol 20 no 6 pp 1328ndash1336 2005

[11] W Jing G Tan and Z Ye ldquoLosses calculation and heat dissipa-tion analysis of high-power three-level convertersrdquoTransactionsof China Electrotechnical Society vol 26 no 2 pp 134ndash140 2011

[12] K Ma Y Yang and F Blaabjerg ldquoTransient modelling ofloss and thermal dynamics in power semiconductor devicesrdquoin Proceedings of the IEEE Energy Conversion Congress andExposition (ECCE rsquo14) pp 5495ndash5501 IEEE 2014

[13] U R Prasanna and A K Rathore ldquoAnalysis design and exper-imental results of a novel soft-switching snubberless current-fed half-bridge front-end converter-based pv inverterrdquo IEEETransactions on Power Electronics vol 28 no 7 pp 3219ndash32302013

[14] X Ruan L Zhou and Y Yan ldquoSoft-switching PWM three-levelconvertersrdquo IEEE Transactions on Power Electronics vol 16 no5 pp 612ndash622 2001

[15] R Li andD Xu ldquoA zero-voltage switching three-phase inverterrdquoIEEE Transactions on Power Electronics vol 29 no 3 pp 1200ndash1210 2014

[16] P Kollensperger R U Lenke S Schroder and R W de Don-cker ldquoDesign of a flexible control platform for soft-switchingmultilevel invertersrdquo IEEE Transactions on Power Electronicsvol 22 no 5 pp 1778ndash1785 2007

[17] G Ortiz H Uemura D Bortis J W Kolar and O ApeldoornldquoModeling of soft-switching losses of IGBTs in high-powerhigh-efficiency dual-active-bridge DCDC convertersrdquo IEEETransactions on Electron Devices vol 60 no 2 pp 587ndash5972013

[18] M W Gekeler ldquoSoft switching three level inverter (S3Linverter)rdquo in Proceedings of the 15th European Conference onPower Electronics and Applications (EPE rsquo13) pp 1ndash10 IEEESeptember 2013

[19] S Munk-Nielsen L N Tutelea and U Jaeger ldquoSimulation withideal switchmodels combinedwithmeasured loss data providesa good estimate of power lossrdquo in Proceedings of the ConferenceRecord of the IEEE Industry Applications Conference vol 5 pp2915ndash2922 Rome Italy October 2000

[20] O S Senturk L Helle S Munk-Nielsen P Rodriguez and RTeodorescu ldquoConverter structure-based power loss and staticthermal modeling of the press-pack IGBT three-level ANPCVSC applied to multi-MW wind turbinesrdquo IEEE Transactionson Industry Applications vol 47 no 6 pp 2505ndash2515 2011

[21] D A B Zambra C Rech F A S Goncalves and J R PinheiroldquoPower losses analysis and cooling system design of threetopologies of multilevel invertersrdquo in Proceedings of the 39thIEEE Annual Power Electronics Specialists Conference (PESCrsquo08) pp 4290ndash4295 IEEE June 2008

[22] W Jing Three-level inverter power loss power devices research[Doctoral dissertation] China University of Mining and Tech-nology Xuzhou China 2011

[23] D Floricau E Floricau and G Gateau ldquoThree-level ActiveNPC converter PWM strategies and loss distributionrdquo inProceedings of the 34th Annual Conference of the IEEE IndustrialElectronics Society (IECON rsquo08) pp 3333ndash3338November 2008

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

Journal of Electrical and Computer Engineering 11

79828

74736

69643

6455

59458

54365

49272

4418

39087

33994

(a)

7561

70822

66035

61247

5646

51673

46885

42098

37311

32523

(b)

79828

79233

78637

78042

77446

76851

76255

7566

75064

74469

(c)

7561

7502

7443

73841

73251

72662

72072

71483

70893

70303

(d)

Figure 16 The temperature distribution 3D map of maternal thermal models before modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

a tool to support the accurate power-loss calculation andthermal analysis using the S3L inverter based on the softswitching control to improve the NPC three-level invertercan get a good result that S3L inverter has the same excellentoutput characteristic with NPC three-level inverter and hasa great advantage in reducing power loss with a 213Wdecline in each half-bridge and 10∘C decline on power devicetemperatureThe thermal stability of three-level inverters canbe enhanced by this improvement

5 Conclusions

The optimized maternal power-loss thermal models of NPCthree-level inverter and S3L inverter were established based

on the optimized power-loss algorithm and a set of generaloptimized power-loss calculation formulas was derived tomodify the total power loss and figure out the modificationpower-loss values Then these values were considered asthermal sources to analyze the maternal thermal modelsThe three-level inverter can be improved by comparing andanalyzing power-loss modification values and experimentresults Based on this principle and methods the fact thatunder the same conditions the power-lossmodification valueof S3L inverter is smaller than that ofNPC three-level inverterby 213W and has a 9∘C advantage is obtained which is only1∘C smaller than the experiment result Experimental resultsvalidate the proposed theoretical calculation and analysis andprove the effectiveness of the improvement

12 Journal of Electrical and Computer Engineering

61678

58801

55924

53047

5017

47293

44416

41539

38662

35785

(a)

49007

4728

45554

43828

42102

40376

38649

36923

35197

33471

(b)

61678

61374

6107

60766

60462

60159

59855

59551

59247

58943

(c)

49007

48824

48642

4846

48277

48095

47913

47731

47548

47366

(d)

Figure 17 The temperature distribution 3D map of maternal thermal models after modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

Appendix

The parameters of double-fed induction motor used in thisexperiment are as follows

rated power 2100KWrated speed 1000 rpmnumber of pole-pairs 3efficiency at full load 97network voltage 690Vnetwork frequency 50Hzstator current 1662A

rotor current 743Acoupling stator Δ rotor Yrotor open voltage 1710Vinertia 94 kgm2stator maximum short-circuit current 4250Arotor maximum short-circuit current 2900Astator resistance R1 0006013Ωstator leakage reactance X1 0045062Ωrotor resistance (equivalent) R2 0004193Ωrotor leakage reactance (equivalent) X2 02298Ω

Journal of Electrical and Computer Engineering 13

NPC3L

(∘C)

603

56

52

48

44

40

36

32

28

246

(a)

S3L

552

51

48

45

42

39

36

33

30

27

247

(∘C )

(b)

NPC3L

(c)

S3L

(d)

Figure 18The experimental temperature results of IGBTrsquos surface and substrate (a) IGBT surface temperature of NPC3L inverter (b) IGBTsurface temperature of S3L inverter (c) IGBT substrate temperature of NPC3L inverter (d) IGBT substrate temperature of S3L inverter

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors would like to thank 2014 Jiangsu ProvinceNatural Science Foundation (BK20140204) the Research andInnovation Program of Postgraduates in Jiangsu Province(CXZZ13 0930) and the Fundamental Research Funds forthe Central Universities (2012LWB73)

References

[1] S Bernet ldquoRecent developments of high power converters forindustry and traction applicationsrdquo IEEE Transactions on PowerElectronics vol 15 no 6 pp 1102ndash1117 2000

[2] P Mao S-J Xie and Z-G Xu ldquoSwitching transients model andloss analysis of IGBTmodulerdquo Proceedings of the Chinese Societyof Electrical Engineering vol 15 article 007 2010

[3] A D Rajapakse AM Gole and P LWilson ldquoElectromagnetictransients simulation models for accurate representation ofswitching losses and thermal performance in power electronic

systemsrdquo IEEE Transactions on Power Delivery vol 20 no 1 pp319ndash327 2005

[4] J Hu J Li J Zou and J Tan ldquoLosses calculation of IGBTmodule and heat dissipation system design of invertersrdquo Trans-actions of China Electrotechnical Society vol 24 no 3 pp 159ndash163 2009

[5] M H Bierhoff and F W Fuchs ldquoSemiconductor losses involtage source and current source IGBT converters based onanalytical derivationrdquo in Proceedings of the IEEE 35th AnnualPower Electronics Specialists Conference (PESC rsquo04) vol 4 pp2836ndash2842 IEEE June 2004

[6] F Krismer and J W Kolar ldquoAccurate power loss modelderivation of a high-current dual active bridge converter foran automotive applicationrdquo IEEE Transactions on IndustrialElectronics vol 57 no 3 pp 881ndash891 2010

[7] Q Chen Q Wang W Jiang and C Hu ldquoAnalysis of switchinglosses in diode-clamped three-level converterrdquo Transactions ofChina Electrotechnical Society vol 23 no 2 pp 68ndash75 2008

[8] J Wang Q Chen W Jiang and C Hu ldquoAnalysis of conductionlosses in neutral-point-clamped three-level inverterrdquo Transac-tions of China Electrotechnical Society vol 3 p 12 2007

[9] T J Kim DW Kang Y H Lee andD S Hyun ldquoThe analysis ofconduction and switching losses inmulti-level inverter systemrdquoin Proceedings of the IEEE 32nd Annual Power ElectronicsSpecialists Conference (PESC rsquo01) pp 1363ndash1368 June 2001

14 Journal of Electrical and Computer Engineering

[10] S Dieckerhoff S Bernet and D Krug ldquoPower loss-orientedevaluation of high voltage IGBTs and multilevel convertersin transformerless traction applicationsrdquo IEEE Transactions onPower Electronics vol 20 no 6 pp 1328ndash1336 2005

[11] W Jing G Tan and Z Ye ldquoLosses calculation and heat dissipa-tion analysis of high-power three-level convertersrdquoTransactionsof China Electrotechnical Society vol 26 no 2 pp 134ndash140 2011

[12] K Ma Y Yang and F Blaabjerg ldquoTransient modelling ofloss and thermal dynamics in power semiconductor devicesrdquoin Proceedings of the IEEE Energy Conversion Congress andExposition (ECCE rsquo14) pp 5495ndash5501 IEEE 2014

[13] U R Prasanna and A K Rathore ldquoAnalysis design and exper-imental results of a novel soft-switching snubberless current-fed half-bridge front-end converter-based pv inverterrdquo IEEETransactions on Power Electronics vol 28 no 7 pp 3219ndash32302013

[14] X Ruan L Zhou and Y Yan ldquoSoft-switching PWM three-levelconvertersrdquo IEEE Transactions on Power Electronics vol 16 no5 pp 612ndash622 2001

[15] R Li andD Xu ldquoA zero-voltage switching three-phase inverterrdquoIEEE Transactions on Power Electronics vol 29 no 3 pp 1200ndash1210 2014

[16] P Kollensperger R U Lenke S Schroder and R W de Don-cker ldquoDesign of a flexible control platform for soft-switchingmultilevel invertersrdquo IEEE Transactions on Power Electronicsvol 22 no 5 pp 1778ndash1785 2007

[17] G Ortiz H Uemura D Bortis J W Kolar and O ApeldoornldquoModeling of soft-switching losses of IGBTs in high-powerhigh-efficiency dual-active-bridge DCDC convertersrdquo IEEETransactions on Electron Devices vol 60 no 2 pp 587ndash5972013

[18] M W Gekeler ldquoSoft switching three level inverter (S3Linverter)rdquo in Proceedings of the 15th European Conference onPower Electronics and Applications (EPE rsquo13) pp 1ndash10 IEEESeptember 2013

[19] S Munk-Nielsen L N Tutelea and U Jaeger ldquoSimulation withideal switchmodels combinedwithmeasured loss data providesa good estimate of power lossrdquo in Proceedings of the ConferenceRecord of the IEEE Industry Applications Conference vol 5 pp2915ndash2922 Rome Italy October 2000

[20] O S Senturk L Helle S Munk-Nielsen P Rodriguez and RTeodorescu ldquoConverter structure-based power loss and staticthermal modeling of the press-pack IGBT three-level ANPCVSC applied to multi-MW wind turbinesrdquo IEEE Transactionson Industry Applications vol 47 no 6 pp 2505ndash2515 2011

[21] D A B Zambra C Rech F A S Goncalves and J R PinheiroldquoPower losses analysis and cooling system design of threetopologies of multilevel invertersrdquo in Proceedings of the 39thIEEE Annual Power Electronics Specialists Conference (PESCrsquo08) pp 4290ndash4295 IEEE June 2008

[22] W Jing Three-level inverter power loss power devices research[Doctoral dissertation] China University of Mining and Tech-nology Xuzhou China 2011

[23] D Floricau E Floricau and G Gateau ldquoThree-level ActiveNPC converter PWM strategies and loss distributionrdquo inProceedings of the 34th Annual Conference of the IEEE IndustrialElectronics Society (IECON rsquo08) pp 3333ndash3338November 2008

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

12 Journal of Electrical and Computer Engineering

61678

58801

55924

53047

5017

47293

44416

41539

38662

35785

(a)

49007

4728

45554

43828

42102

40376

38649

36923

35197

33471

(b)

61678

61374

6107

60766

60462

60159

59855

59551

59247

58943

(c)

49007

48824

48642

4846

48277

48095

47913

47731

47548

47366

(d)

Figure 17 The temperature distribution 3D map of maternal thermal models after modification (a) One heat sink temperature chart ofNPC3L inverter (b) one heat sink temperature chart of S3L inverter (c) one substrate temperature chart of NPC3L inverter (d) one substratetemperature chart of S3L inverter

Appendix

The parameters of double-fed induction motor used in thisexperiment are as follows

rated power 2100KWrated speed 1000 rpmnumber of pole-pairs 3efficiency at full load 97network voltage 690Vnetwork frequency 50Hzstator current 1662A

rotor current 743Acoupling stator Δ rotor Yrotor open voltage 1710Vinertia 94 kgm2stator maximum short-circuit current 4250Arotor maximum short-circuit current 2900Astator resistance R1 0006013Ωstator leakage reactance X1 0045062Ωrotor resistance (equivalent) R2 0004193Ωrotor leakage reactance (equivalent) X2 02298Ω

Journal of Electrical and Computer Engineering 13

NPC3L

(∘C)

603

56

52

48

44

40

36

32

28

246

(a)

S3L

552

51

48

45

42

39

36

33

30

27

247

(∘C )

(b)

NPC3L

(c)

S3L

(d)

Figure 18The experimental temperature results of IGBTrsquos surface and substrate (a) IGBT surface temperature of NPC3L inverter (b) IGBTsurface temperature of S3L inverter (c) IGBT substrate temperature of NPC3L inverter (d) IGBT substrate temperature of S3L inverter

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors would like to thank 2014 Jiangsu ProvinceNatural Science Foundation (BK20140204) the Research andInnovation Program of Postgraduates in Jiangsu Province(CXZZ13 0930) and the Fundamental Research Funds forthe Central Universities (2012LWB73)

References

[1] S Bernet ldquoRecent developments of high power converters forindustry and traction applicationsrdquo IEEE Transactions on PowerElectronics vol 15 no 6 pp 1102ndash1117 2000

[2] P Mao S-J Xie and Z-G Xu ldquoSwitching transients model andloss analysis of IGBTmodulerdquo Proceedings of the Chinese Societyof Electrical Engineering vol 15 article 007 2010

[3] A D Rajapakse AM Gole and P LWilson ldquoElectromagnetictransients simulation models for accurate representation ofswitching losses and thermal performance in power electronic

systemsrdquo IEEE Transactions on Power Delivery vol 20 no 1 pp319ndash327 2005

[4] J Hu J Li J Zou and J Tan ldquoLosses calculation of IGBTmodule and heat dissipation system design of invertersrdquo Trans-actions of China Electrotechnical Society vol 24 no 3 pp 159ndash163 2009

[5] M H Bierhoff and F W Fuchs ldquoSemiconductor losses involtage source and current source IGBT converters based onanalytical derivationrdquo in Proceedings of the IEEE 35th AnnualPower Electronics Specialists Conference (PESC rsquo04) vol 4 pp2836ndash2842 IEEE June 2004

[6] F Krismer and J W Kolar ldquoAccurate power loss modelderivation of a high-current dual active bridge converter foran automotive applicationrdquo IEEE Transactions on IndustrialElectronics vol 57 no 3 pp 881ndash891 2010

[7] Q Chen Q Wang W Jiang and C Hu ldquoAnalysis of switchinglosses in diode-clamped three-level converterrdquo Transactions ofChina Electrotechnical Society vol 23 no 2 pp 68ndash75 2008

[8] J Wang Q Chen W Jiang and C Hu ldquoAnalysis of conductionlosses in neutral-point-clamped three-level inverterrdquo Transac-tions of China Electrotechnical Society vol 3 p 12 2007

[9] T J Kim DW Kang Y H Lee andD S Hyun ldquoThe analysis ofconduction and switching losses inmulti-level inverter systemrdquoin Proceedings of the IEEE 32nd Annual Power ElectronicsSpecialists Conference (PESC rsquo01) pp 1363ndash1368 June 2001

14 Journal of Electrical and Computer Engineering

[10] S Dieckerhoff S Bernet and D Krug ldquoPower loss-orientedevaluation of high voltage IGBTs and multilevel convertersin transformerless traction applicationsrdquo IEEE Transactions onPower Electronics vol 20 no 6 pp 1328ndash1336 2005

[11] W Jing G Tan and Z Ye ldquoLosses calculation and heat dissipa-tion analysis of high-power three-level convertersrdquoTransactionsof China Electrotechnical Society vol 26 no 2 pp 134ndash140 2011

[12] K Ma Y Yang and F Blaabjerg ldquoTransient modelling ofloss and thermal dynamics in power semiconductor devicesrdquoin Proceedings of the IEEE Energy Conversion Congress andExposition (ECCE rsquo14) pp 5495ndash5501 IEEE 2014

[13] U R Prasanna and A K Rathore ldquoAnalysis design and exper-imental results of a novel soft-switching snubberless current-fed half-bridge front-end converter-based pv inverterrdquo IEEETransactions on Power Electronics vol 28 no 7 pp 3219ndash32302013

[14] X Ruan L Zhou and Y Yan ldquoSoft-switching PWM three-levelconvertersrdquo IEEE Transactions on Power Electronics vol 16 no5 pp 612ndash622 2001

[15] R Li andD Xu ldquoA zero-voltage switching three-phase inverterrdquoIEEE Transactions on Power Electronics vol 29 no 3 pp 1200ndash1210 2014

[16] P Kollensperger R U Lenke S Schroder and R W de Don-cker ldquoDesign of a flexible control platform for soft-switchingmultilevel invertersrdquo IEEE Transactions on Power Electronicsvol 22 no 5 pp 1778ndash1785 2007

[17] G Ortiz H Uemura D Bortis J W Kolar and O ApeldoornldquoModeling of soft-switching losses of IGBTs in high-powerhigh-efficiency dual-active-bridge DCDC convertersrdquo IEEETransactions on Electron Devices vol 60 no 2 pp 587ndash5972013

[18] M W Gekeler ldquoSoft switching three level inverter (S3Linverter)rdquo in Proceedings of the 15th European Conference onPower Electronics and Applications (EPE rsquo13) pp 1ndash10 IEEESeptember 2013

[19] S Munk-Nielsen L N Tutelea and U Jaeger ldquoSimulation withideal switchmodels combinedwithmeasured loss data providesa good estimate of power lossrdquo in Proceedings of the ConferenceRecord of the IEEE Industry Applications Conference vol 5 pp2915ndash2922 Rome Italy October 2000

[20] O S Senturk L Helle S Munk-Nielsen P Rodriguez and RTeodorescu ldquoConverter structure-based power loss and staticthermal modeling of the press-pack IGBT three-level ANPCVSC applied to multi-MW wind turbinesrdquo IEEE Transactionson Industry Applications vol 47 no 6 pp 2505ndash2515 2011

[21] D A B Zambra C Rech F A S Goncalves and J R PinheiroldquoPower losses analysis and cooling system design of threetopologies of multilevel invertersrdquo in Proceedings of the 39thIEEE Annual Power Electronics Specialists Conference (PESCrsquo08) pp 4290ndash4295 IEEE June 2008

[22] W Jing Three-level inverter power loss power devices research[Doctoral dissertation] China University of Mining and Tech-nology Xuzhou China 2011

[23] D Floricau E Floricau and G Gateau ldquoThree-level ActiveNPC converter PWM strategies and loss distributionrdquo inProceedings of the 34th Annual Conference of the IEEE IndustrialElectronics Society (IECON rsquo08) pp 3333ndash3338November 2008

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

Journal of Electrical and Computer Engineering 13

NPC3L

(∘C)

603

56

52

48

44

40

36

32

28

246

(a)

S3L

552

51

48

45

42

39

36

33

30

27

247

(∘C )

(b)

NPC3L

(c)

S3L

(d)

Figure 18The experimental temperature results of IGBTrsquos surface and substrate (a) IGBT surface temperature of NPC3L inverter (b) IGBTsurface temperature of S3L inverter (c) IGBT substrate temperature of NPC3L inverter (d) IGBT substrate temperature of S3L inverter

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors would like to thank 2014 Jiangsu ProvinceNatural Science Foundation (BK20140204) the Research andInnovation Program of Postgraduates in Jiangsu Province(CXZZ13 0930) and the Fundamental Research Funds forthe Central Universities (2012LWB73)

References

[1] S Bernet ldquoRecent developments of high power converters forindustry and traction applicationsrdquo IEEE Transactions on PowerElectronics vol 15 no 6 pp 1102ndash1117 2000

[2] P Mao S-J Xie and Z-G Xu ldquoSwitching transients model andloss analysis of IGBTmodulerdquo Proceedings of the Chinese Societyof Electrical Engineering vol 15 article 007 2010

[3] A D Rajapakse AM Gole and P LWilson ldquoElectromagnetictransients simulation models for accurate representation ofswitching losses and thermal performance in power electronic

systemsrdquo IEEE Transactions on Power Delivery vol 20 no 1 pp319ndash327 2005

[4] J Hu J Li J Zou and J Tan ldquoLosses calculation of IGBTmodule and heat dissipation system design of invertersrdquo Trans-actions of China Electrotechnical Society vol 24 no 3 pp 159ndash163 2009

[5] M H Bierhoff and F W Fuchs ldquoSemiconductor losses involtage source and current source IGBT converters based onanalytical derivationrdquo in Proceedings of the IEEE 35th AnnualPower Electronics Specialists Conference (PESC rsquo04) vol 4 pp2836ndash2842 IEEE June 2004

[6] F Krismer and J W Kolar ldquoAccurate power loss modelderivation of a high-current dual active bridge converter foran automotive applicationrdquo IEEE Transactions on IndustrialElectronics vol 57 no 3 pp 881ndash891 2010

[7] Q Chen Q Wang W Jiang and C Hu ldquoAnalysis of switchinglosses in diode-clamped three-level converterrdquo Transactions ofChina Electrotechnical Society vol 23 no 2 pp 68ndash75 2008

[8] J Wang Q Chen W Jiang and C Hu ldquoAnalysis of conductionlosses in neutral-point-clamped three-level inverterrdquo Transac-tions of China Electrotechnical Society vol 3 p 12 2007

[9] T J Kim DW Kang Y H Lee andD S Hyun ldquoThe analysis ofconduction and switching losses inmulti-level inverter systemrdquoin Proceedings of the IEEE 32nd Annual Power ElectronicsSpecialists Conference (PESC rsquo01) pp 1363ndash1368 June 2001

14 Journal of Electrical and Computer Engineering

[10] S Dieckerhoff S Bernet and D Krug ldquoPower loss-orientedevaluation of high voltage IGBTs and multilevel convertersin transformerless traction applicationsrdquo IEEE Transactions onPower Electronics vol 20 no 6 pp 1328ndash1336 2005

[11] W Jing G Tan and Z Ye ldquoLosses calculation and heat dissipa-tion analysis of high-power three-level convertersrdquoTransactionsof China Electrotechnical Society vol 26 no 2 pp 134ndash140 2011

[12] K Ma Y Yang and F Blaabjerg ldquoTransient modelling ofloss and thermal dynamics in power semiconductor devicesrdquoin Proceedings of the IEEE Energy Conversion Congress andExposition (ECCE rsquo14) pp 5495ndash5501 IEEE 2014

[13] U R Prasanna and A K Rathore ldquoAnalysis design and exper-imental results of a novel soft-switching snubberless current-fed half-bridge front-end converter-based pv inverterrdquo IEEETransactions on Power Electronics vol 28 no 7 pp 3219ndash32302013

[14] X Ruan L Zhou and Y Yan ldquoSoft-switching PWM three-levelconvertersrdquo IEEE Transactions on Power Electronics vol 16 no5 pp 612ndash622 2001

[15] R Li andD Xu ldquoA zero-voltage switching three-phase inverterrdquoIEEE Transactions on Power Electronics vol 29 no 3 pp 1200ndash1210 2014

[16] P Kollensperger R U Lenke S Schroder and R W de Don-cker ldquoDesign of a flexible control platform for soft-switchingmultilevel invertersrdquo IEEE Transactions on Power Electronicsvol 22 no 5 pp 1778ndash1785 2007

[17] G Ortiz H Uemura D Bortis J W Kolar and O ApeldoornldquoModeling of soft-switching losses of IGBTs in high-powerhigh-efficiency dual-active-bridge DCDC convertersrdquo IEEETransactions on Electron Devices vol 60 no 2 pp 587ndash5972013

[18] M W Gekeler ldquoSoft switching three level inverter (S3Linverter)rdquo in Proceedings of the 15th European Conference onPower Electronics and Applications (EPE rsquo13) pp 1ndash10 IEEESeptember 2013

[19] S Munk-Nielsen L N Tutelea and U Jaeger ldquoSimulation withideal switchmodels combinedwithmeasured loss data providesa good estimate of power lossrdquo in Proceedings of the ConferenceRecord of the IEEE Industry Applications Conference vol 5 pp2915ndash2922 Rome Italy October 2000

[20] O S Senturk L Helle S Munk-Nielsen P Rodriguez and RTeodorescu ldquoConverter structure-based power loss and staticthermal modeling of the press-pack IGBT three-level ANPCVSC applied to multi-MW wind turbinesrdquo IEEE Transactionson Industry Applications vol 47 no 6 pp 2505ndash2515 2011

[21] D A B Zambra C Rech F A S Goncalves and J R PinheiroldquoPower losses analysis and cooling system design of threetopologies of multilevel invertersrdquo in Proceedings of the 39thIEEE Annual Power Electronics Specialists Conference (PESCrsquo08) pp 4290ndash4295 IEEE June 2008

[22] W Jing Three-level inverter power loss power devices research[Doctoral dissertation] China University of Mining and Tech-nology Xuzhou China 2011

[23] D Floricau E Floricau and G Gateau ldquoThree-level ActiveNPC converter PWM strategies and loss distributionrdquo inProceedings of the 34th Annual Conference of the IEEE IndustrialElectronics Society (IECON rsquo08) pp 3333ndash3338November 2008

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

14 Journal of Electrical and Computer Engineering

[10] S Dieckerhoff S Bernet and D Krug ldquoPower loss-orientedevaluation of high voltage IGBTs and multilevel convertersin transformerless traction applicationsrdquo IEEE Transactions onPower Electronics vol 20 no 6 pp 1328ndash1336 2005

[11] W Jing G Tan and Z Ye ldquoLosses calculation and heat dissipa-tion analysis of high-power three-level convertersrdquoTransactionsof China Electrotechnical Society vol 26 no 2 pp 134ndash140 2011

[12] K Ma Y Yang and F Blaabjerg ldquoTransient modelling ofloss and thermal dynamics in power semiconductor devicesrdquoin Proceedings of the IEEE Energy Conversion Congress andExposition (ECCE rsquo14) pp 5495ndash5501 IEEE 2014

[13] U R Prasanna and A K Rathore ldquoAnalysis design and exper-imental results of a novel soft-switching snubberless current-fed half-bridge front-end converter-based pv inverterrdquo IEEETransactions on Power Electronics vol 28 no 7 pp 3219ndash32302013

[14] X Ruan L Zhou and Y Yan ldquoSoft-switching PWM three-levelconvertersrdquo IEEE Transactions on Power Electronics vol 16 no5 pp 612ndash622 2001

[15] R Li andD Xu ldquoA zero-voltage switching three-phase inverterrdquoIEEE Transactions on Power Electronics vol 29 no 3 pp 1200ndash1210 2014

[16] P Kollensperger R U Lenke S Schroder and R W de Don-cker ldquoDesign of a flexible control platform for soft-switchingmultilevel invertersrdquo IEEE Transactions on Power Electronicsvol 22 no 5 pp 1778ndash1785 2007

[17] G Ortiz H Uemura D Bortis J W Kolar and O ApeldoornldquoModeling of soft-switching losses of IGBTs in high-powerhigh-efficiency dual-active-bridge DCDC convertersrdquo IEEETransactions on Electron Devices vol 60 no 2 pp 587ndash5972013

[18] M W Gekeler ldquoSoft switching three level inverter (S3Linverter)rdquo in Proceedings of the 15th European Conference onPower Electronics and Applications (EPE rsquo13) pp 1ndash10 IEEESeptember 2013

[19] S Munk-Nielsen L N Tutelea and U Jaeger ldquoSimulation withideal switchmodels combinedwithmeasured loss data providesa good estimate of power lossrdquo in Proceedings of the ConferenceRecord of the IEEE Industry Applications Conference vol 5 pp2915ndash2922 Rome Italy October 2000

[20] O S Senturk L Helle S Munk-Nielsen P Rodriguez and RTeodorescu ldquoConverter structure-based power loss and staticthermal modeling of the press-pack IGBT three-level ANPCVSC applied to multi-MW wind turbinesrdquo IEEE Transactionson Industry Applications vol 47 no 6 pp 2505ndash2515 2011

[21] D A B Zambra C Rech F A S Goncalves and J R PinheiroldquoPower losses analysis and cooling system design of threetopologies of multilevel invertersrdquo in Proceedings of the 39thIEEE Annual Power Electronics Specialists Conference (PESCrsquo08) pp 4290ndash4295 IEEE June 2008

[22] W Jing Three-level inverter power loss power devices research[Doctoral dissertation] China University of Mining and Tech-nology Xuzhou China 2011

[23] D Floricau E Floricau and G Gateau ldquoThree-level ActiveNPC converter PWM strategies and loss distributionrdquo inProceedings of the 34th Annual Conference of the IEEE IndustrialElectronics Society (IECON rsquo08) pp 3333ndash3338November 2008

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Active and Passive Electronic Components

Control Scienceand Engineering

Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

International Journal of

RotatingMachinery

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Shock and Vibration

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawi Publishing Corporation httpwwwhindawicom

Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Navigation and Observation

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

DistributedSensor Networks

International Journal of