A

Tzfm©B

K

1

gtspiaPse(

2B

Available online at www.sciencedirect.com

ScienceDirect

Journal of Electrical Systems and Information Technology 3 (2016) 81–93

A novel topology and control strategy for a soft-switchedsingle-phase grid connected inverter

Majid Pakdel ∗, Saeid JalilzadehElectronical Engineering Department, University of Zanjan, Zanjan, Iran

Received 11 May 2015; accepted 11 November 2015Available online 18 March 2016

bstract

This paper represents a novel soft switched PWM inverter topology and a control strategy using switching flow graph theory.he switching flow graph is a unique graphical model for analysis of PWM switching converters. All power switches operate atero-voltage-switching (ZVS) turn-ON and turn-OFF conditions as a result, commutations of power switches have lower lossesurthermore higher efficiency is reached. The proposed approach is analyzed theoretically and demonstrated experimentally byaking a testing prototype.

2016 Electronics Research Institute (ERI). Production and hosting by Elsevier B.V. This is an open access article under the CCY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

eywords: Renewable energy; DC–AC power converters; Zero-voltage-switching; Switching flow graph theory; Soft-switching

. Introduction

Solar and wind energies are the most significant sources of renewable energies available in the world. Furthermore,rid connected inverters are increasing and developing quickly and their performance are improved day to day. Dueo the increased electrical energy demand globally the mentioned renewable energy resources are the most effectiveolutions. However, the most significant disadvantage of such renewable energy resources is that their produced outputower is discrete and unregulated since some techniques such as maximum power point tracking (MPPT) methods used to get constant output voltage and power out of solar arrays or wind turbines. Various inverter topologiesnd control strategies have been proposed for grid connected inverters in literatures (Pierquet and Perreault, 2013;rasanna and Rathore, 2013; Batzel and Adams, 2013). Soft switching techniques are used extensively to reduce

witching losses associated with power switch commutations (Li and Wolfs, 2007; Jain and Agarwal, 2008; Todorovict al., 2008; Rathore et al., 2008). Switching converters are pulsed converters and have nonlinear dynamic behaviorsSmedley, 1991). Linear feedback control structures are significantly used to control nonlinear systems. If a system

∗ Corresponding author. Tel.: +98 9141248004.E-mail addresses: pakdel@mail.ru, pakdel@znu.ac.ir (M. Pakdel).Peer review under responsibility of Electronics Research Institute (ERI).

http://dx.doi.org/10.1016/j.jesit.2015.11.011314-7172/© 2016 Electronics Research Institute (ERI). Production and hosting by Elsevier B.V. This is an open access article under the CCY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

82 M. Pakdel, S. Jalilzadeh / Journal of Electrical Systems and Information Technology 3 (2016) 81–93

is more complicated the conventional linear control theory approach may not be useful. In this paper a novel uniquenonlinear modeling method, the switching flow graph technique, is developed and applied to demonstrate the nonlineardynamic behavior of power switching converters. This technique utilizes some concepts from state space averagingmethod as well as extension of linear circuit flow graph theory (Smedley, 1991). A switching converter is equivalentto ON-circuit when the switch is in the ON state and it is identical to OFF-circuit when the switch is in the OFFstate. The switching flow graph is extracted by combining the flow graphs of the ON-circuit and the OFF-circuit. Theswitching flow graph decomposes a system to its basic components and clearly evaluates the effects of input parameterchanges on system outputs. Switching flow graphs can be used to extract transfer functions between arbitrary points inthe power switching converter. The switching flow graph can easily be concluded from the electrical system topology.The signals are represented by nodes which are depicted by small signals in the switching flow graph. The nodes areconnected by branches which are arrows moreover the signals flow only in the arrow direction furthermore each arrowhas a gain illustrated next to the arrow. The signal at a node is the sum of all signals entering to the node. Nodeswith just outputs and no inputs are called source nodes and represent independent variables. However, sink nodes,nodes with only inputs and no outputs show dependent variables on the other hand mixed nodes have both inputs andoutputs.

The procedure for building a switching flow graph is very simple. First, all electrical variables of elements i.e.voltages and currents of all electrical elements in power switching converter are assigned nodes in the switching flowgraph. Next, the nodes are interconnected with arrows according to electrical rules between voltages and currents ofall elements. Once all the nodes are conveniently connected the switching flow graph will be completed. The powerconverter presented in this paper implements a novel type of grid-connected topology as well as a novel nonlinear

control strategy using switching flow graph theory. The proposed topology has higher efficiency in addition it operatesunder continuous constant output power condition furthermore that provides zero-voltage switching (ZVS) capabilityfor all power switches.

Fig. 1. Proposed grid-tied inverter: (a) block diagram and (b) circuit schematic.

2

hspt

M. Pakdel, S. Jalilzadeh / Journal of Electrical Systems and Information Technology 3 (2016) 81–93 83

. Proposed topology and control strategy

The block diagram and circuit schematic in Fig. 1 depict four functional blocks of proposed tie-grid inverter: PWMalf bridge inverter, filter, cycloconverter and output transformer. Each block is connected in series electrically. This

eries connected configuration may seem to have heavy conduction losses however by using zero voltage switching ofower switches the losses associated with this impact can be greatly declined (Henze et al., 1988). This study showshat this approach has high efficiency furthermore in contrast with traditional converters the algorithms for maximum

Fig. 2. Switching flow graph of the proposed grid-tied inverter.

Fig. 3. Switching functions.

Fig. 4. Models of the switching branches.

84 M. Pakdel, S. Jalilzadeh / Journal of Electrical Systems and Information Technology 3 (2016) 81–93

Fig. 5. Operation modes of the proposed grid-tied inverter.

M. Pakdel, S. Jalilzadeh / Journal of Electrical Systems and Information Technology 3 (2016) 81–93 85

power point tracking (MPPT) in photovoltaic converts (Esram and Chapman, 2007), grid synchronization (Blaabjerget al., 2006) and islanding detection methods (Ye et al., 2004) can extensively be used within this proposed topology.The switching flow graph is extended to study the dynamic behavior of the proposed grid tied inverter. Suppose theswitching converter operates at a frequency as follows:

fs(t) = 1

Ts(t)(1)

Switches S1, S3 and S6 are ON and switches S2, S4 and S5 are OFF when 0 < t< TON moreover switches S2, S4 andS5 are ON and switches S1, S3 and S6 are OFF when TON < t< Ts. It is clear that a change in circuit topology occurswithin each period since the circuit configuration is periodically changed from the ON-circuit to the OFF-circuit. Bothsub-circuits are linear however the switching converter is a nonlinear circuit due to the periodic topology changes.During the time when 0 < t < TON, the converter is switched to the ON-circuit and during the time when TON < t< Ts,the converter is switched to the OFF-circuit. The switching flow graph of the proposed grid-tied inverter is shown inFig. 2.

The switching functions k and k̄ are shown in Fig. 3. The k function has a value “1” when 0 < t < TON and has avalue “0” when TON < t < Ts. The k function has a value of 1 − k. With these two switching functions two switchingcircuits i.e. ON and OFF circuits can be topologically merged as depicted in Fig. 2.

The effective signal at the k-branch output is equal to the average value over a switch cycle:

y(t) = 1

Ts(t)

∫ TON(t)

0x(t)dt ≈ x(t)

1

Ts(t)

∫ TON(t)

0dt = x(t)

TON(t)

Ts(t)= x(t)d(t) (2)

Fig. 6. Proposed grid-tied inverter simulation using PSIM.

86 M. Pakdel, S. Jalilzadeh / Journal of Electrical Systems and Information Technology 3 (2016) 81–93

Fig. 7. Control system extracted from switching flow graph.

Fig. 8. Zero voltage PWM switching pulses generator.

Fig. 9. Transformer primary voltage waveform.

M. Pakdel, S. Jalilzadeh / Journal of Electrical Systems and Information Technology 3 (2016) 81–93 87

wt

wr

Fig. 10. Current iL1 waveform.

here x(t) and y(t) are the input and output nodes in the k-branch of switching flow graph respectively. Similarly forhe k̄-branch, the input signal x(t) and output signal y(t) have the following relation:

y(t) = 1

Ts(t)

∫ Ts(t)

TON(t)x(t)dt ≈ x(t)

1

Ts(t)

∫ Ts(t)

TON(t)x(t)dt = x(t)

TOFF(t)

Ts(t)= x(t)d′(t) (3)

here d(t) and d′(t) are the averages of the switch functions k(t) and k̄(t) respectively, moreover they represent the dutyatio of the switches and we have the following equations:

d(t) = TON(t)

Ts(t)(4)

d′(t) = TOFF(t)

Ts(t)(5)

d′(t) = 1 − d(t) (6)

Fig. 11. PWM switching gate pulses of power devices.

88 M. Pakdel, S. Jalilzadeh / Journal of Electrical Systems and Information Technology 3 (2016) 81–93

Fig. 12. Output control signal of switching flow graph.

Therefore, the models of the switching branches are represented by a single multiplier as shown in Fig. 4. The operationmodes of the proposed grid tied inverter are shown in Fig. 5. In operation mode I, the power switches S1, S3 and S6 areON switches S2, S4 and S5 are OFF and current flow direction is Vdc–S1–L1–S3–C4–S6–C5–L2–C2. Furthermore wehave the following equation:

L1di

dt+ 1

C4

∫ t1

0i dt + 1

C5

∫ t1

0i dt + L2

di

dt+ 1

C2

∫ t1

0i dt = Vdc (7)

In operation mode II, all power switches S1, S2, S3, S4, S5 and S6 are OFF and the current flows through reverse diodesof switches S2, S4 and S5 moreover D1, L1, C3, C5 and L2. In addition the following equation is used for describingsystem behavior:

L1di

dt+ 1

C3

∫ t2

t1

i dt + L2di

dt+ 1

C5

∫ t2

t1

i dt = −Vdc (8)

Fig. 13. Zero voltage switching of S1.

M. Pakdel, S. Jalilzadeh / Journal of Electrical Systems and Information Technology 3 (2016) 81–93 89

Ic

Io

Fig. 14. Zero voltage switching of S3.

n operation mode III, the power switches S1, S3and S6 are OFF and switches S2, S4 and S5 are ON. Furthermoreurrent flow direction is Vdc–C1–L2–S4–C3–S5–L1–S2 as well as the following equation can be written:

L1di

dt+ 1

C3

∫ t3

t2

i dt + 1

C5

∫ t3

t2

i dt + L2di

dt+ 1

C2

∫ t3

t2

i dt = Vdc (9)

n operation mode IV, all power switches S1, S2, S3, S4, S5 and S6 are OFF and the current flows through reverse diodesf switches S1, S3 and S6 moreover D2, L2, C5, C4 and L1. Furthermore the following equation is obtained:

L1di + 1

∫ t4

i dt + L2di + 1

∫ t4

i dt = −Vdc (10)

dt C4 t3 dt C5 t3

Fig. 15. Zero voltage switching of S5.

90 M. Pakdel, S. Jalilzadeh / Journal of Electrical Systems and Information Technology 3 (2016) 81–93

3. Simulation results

Simulation of topology and control strategy for the proposed grid-tied inverter is implemented using PSIM softwareand is shown in Fig. 6. The following parameters and values have been used in simulation study:

L1 = L2 = 100 �H, C1 = C2 = C6 = C7 = 2.2 �F, C3 = C4 = 1 �F,

C5 = 3.3 �F, fs = 1 kHz, Cout = 1 �F

Also, the output transformer has the following parameters:

Lm(magnetizing inductance) = 0.1 H, NP = 50 and NS = 500

The proposed control strategy includes two segments. The first segment is extracted from switching flow graph whichis depicted in Fig. 2 and in more details is shown in Fig. 7 moreover the second segment as shown clearly in Fig. 8 isused for zero voltage PWM switching pulses generation.

The simulated signals as measured are evaluated and analyzed the soft switching operation of power switchesfurthermore its ability to provide pure sine wave signal which is synchronous with grid network is investigated withmore details. Transformer primary voltage waveform is shown in Fig. 9. The current iL1 waveform is illustrated inFig. 10.

PWM zero voltage switching pulses of all power devices are depicted in Fig. 11. Output control signal of switchingflow graph is shown in Fig. 12 furthermore drain to source voltage, drain current and switching gate pulses of power

Fig. 16. A photograph of the proposed grid-tied inverter.

Fig. 17. PWM gate signal generator circuit.

M. Pakdel, S. Jalilzadeh / Journal of Electrical Systems and Information Technology 3 (2016) 81–93 91

switches S1, S3 and S5 are illustrated in Figs. 13, 14 and 15respectively. As shown in Figs. 13–15, the zero voltageswitching conditions are satisfied therefore the switching losses are declined significantly.

4. Experimental results

To illustrate the performance and functionality of the proposed grid-tied inverter as described in this paper, aprototype platform has been designed furthermore a photograph of the proposed converter can be seen in Fig. 16.Also the PWM gate signal generator circuit is shown in Fig. 17. The following circuit elements and parameters havebeen used in the proposed grid-tied inverter prototype:

Input DC voltage: 12 VOutput AC voltage: 220

√2 sin[2�(50)t]

Output power: 500 W maximumInductors L1 = L2 = 100 �HCapacitors C1 = C2 = C6 = C7 = 2.2 �F, C3 = C4 = 1 �F and C5 = 3.3 �F

Fig. 18. Real waveforms of PWM gate signals.

Fig. 19. Voltage waveform across capacitors C3 and C4.

92 M. Pakdel, S. Jalilzadeh / Journal of Electrical Systems and Information Technology 3 (2016) 81–93

100 150 200 250 300 350 400 450 50076

78

80

82

84

86

88

90

92

Output Power (W)

Eff

icie

ncy (

%)

Proposed Soft-switched

Grid-tied Inverter

Hard Switching

PWM Inverter

Fig. 20. Efficiency comparison of the proposed soft-switched grid-tied inverter and traditional hard switching PWM inverter.

Output transformer turns ratio: 25Power switches: IRF840Power diodes: 1N4007Measured waveforms of PWM gate signals for power switches driving circuits using a digital oscilloscope is shown

in Fig. 18 also voltage waveform across capacitors C3 and C4 illustrated in Fig. 19.

The efficiency comparison diagram of the proposed soft switched grid connected inverter and traditional hardswitching inverter is shown in Fig. 20 which shows that the proposed converter has 92% efficiency at rated power.

5. Conclusions

The grid-tied inverter design and implementation presented in this paper has demonstrated a new topology andcontrol strategy using switching flow graph theory. In addition, the proposed inverter maintains soft-switching ZVSturn ON and turn OFF of all power devices moreover simulation results using PSIM software have been presented tovalidate the proposed analysis and design. The presented prototype verifies of the functionality and performance ofthe design and analysis. Experimental results show the accuracy of the analysis and high performance of the proposedtopology and control strategy.

References

Batzel, T.D., Adams, K., 2013. Variable timing control for ARCP voltage source inverters operating at low dc voltage. Int. J. Mod. Eng. 13 (12).Blaabjerg, F., Teodorescu, R., Liserre, M., Timbus, A.V., 2006. Overview of control and grid synchronization for distributed power generation

systems. IEEE Trans. Ind. Electron. 53 (October (5)), 1398–1409.Esram, T., Chapman, P.L., 2007. Comparison of photovoltaic array maximum power point tracking techniques. IEEE Trans. Energy Convers. 22

(June (2)), 439–449.Henze, C.P., Martin, H.C., Parsley, D.W., 1988. Zero-voltage switching in high frequency power converters using pulse width modulation. In: Proc.

3rd Annu. IEEE Appl. Power Electron. Conf. Expo., February, pp. 33–40.Jain, S., Agarwal, V., 2008. An integrated hybrid power supply for distributed generation applications fed by nonconventional energy sources. IEEE

Trans. Energy Convers. 23 (June (2)), 622–631.Li, Q., Wolfs, P., 2007. A current fed two-inductor boost converter with an integrated magnetic structure and passive lossless snubbers for photovoltaic

module integrated converter applications. IEEE Trans. Power Electron. 22 (January (1)), 309–321.

Pierquet, B.J., Perreault, D.J., 2013. A single phase photovoltaic inverter topology with a series-connected energy buffer. IEEE Trans. Power Electron.

28 (October (10)).Prasanna, U.R., Rathore, A.K., 2013. Analysis, design and experimental results of a novel soft switching snubberless current-fed half-bridge front-end

converter-based PV inverter. IEEE Trans. Power Electron. 28 (July (7)).

R

ST

Y

M. Pakdel, S. Jalilzadeh / Journal of Electrical Systems and Information Technology 3 (2016) 81–93 93

athore, A.K., Bhat, A.K.S., Oruganti, R., 2008. A comparison of soft-switched dc–dc converters for fuel cell to utility interface application. IEEJTrans. Ind. Appl. 128 (4), 450–458.

medley, K.M., (PhD thesis) 1991. Control Art of Switching Converters. California Institute of Technology.

odorovic, M.H., Palma, L., Enjeti, P., 2008. Design of a wide input range dc–dc converter with a robust power control scheme suitable for fuel cell

power conversion. IEEE Trans. Ind. Electron. 55 (March (3)), 1247–1255.e, Z., Kolwalkar, A., Zhang, Y., Du, P., Walling, R., 2004. Evaluation of anti-islanding schemes based on nondetection zone concept. IEEE Trans.

Power Electron. 19 (September (5)), 1171–1176.