condition for real-time measurement of power of unsteady ......3/21/2018 kazushi sanada 6....

15
3/21/2018 Kazushi SANADA Kazushi SANADA Yokohama National University, Japan Condition for Real-time Measurement of Power of Unsteady Fluid Flow in a Pipe by Kalman Filter 1 p up p mid p down B pd A pd 1 s c + - g + + + + p mid x c f q mid P mid L 1 L 2

Upload: others

Post on 07-Feb-2021

1 views

Category:

Documents


0 download

TRANSCRIPT

  • 3/21/2018Kazushi SANADA

    Kazushi SANADAYokohama National University, Japan

    Condition for Real-time Measurement of Power of Unsteady Fluid Flow in a Pipe by Kalman Filter

    1

    pup pmid pdown

    Bpd

    Apd

    1s c

    + -

    g

    ++

    ++

    pmid

    x

    cf

    qmid

    Pmid

    L1 L2

  • 1

    2

    3

    4

    5

    6

    3/21/2018Kazushi SANADA

    Introduction

    Kalman filter theory

    Off-line estimation of unsteady flow rate by Kalman filter

    Plant model of Kalman filter: Pipeline dynamics model

    Real-time implementation of Kalman filter

    Conclusion

    2

  • 3/21/2018Kazushi SANADA

    1. Introduction

    3

    Electromagnetic flowmeter

    ○Direct measurement in real-time

    ×sensor may disturb flow×sensor is influenced by velocity distribution×time-special distribution is not obtained

    Coriolis flow meter

    Laser Doppler flow velocimeter

    http://www.kanomax.co.jp/fsmart.html

    http://www.keyence.co.jp/atsuryoku/ryuryou/fd_s/

    http://www.keyence.co.jp/atsuryoku/ryuryou/fd_m/

    http://www.fujielectric.co.jp/products/instruments/products/flow_ultra/FSV.html

    Ultrasonic flowmeterConductivity

    U-shape tube

    Velocity distribution Pin-point Velocity

  • 3/21/2018Kazushi SANADA

    1. Introduction

    • Kalman filter is proposed to estimate unsteady fluid flow in a pipe using a model of pipeline dynamics.

    • The merit of this method is to use only three pressure sensors without flow sensor which may disturb flow in a pipe.

    • The aim of this research is to develop a real-time indirect measurement system of unsteady flow rate and fluid power in a pipe.

    4

  • 3/21/2018Kazushi SANADA

    2. Kalman filter theory

    5

    Kalman filter estimates the optimal state variables of a system which is disturbed by noise.

    1 1

    1 , and

    .

    1 1 , andPriori estimation

    Posteriori estimation

    Kalman gain

    Riccati equation

    Variances

    ,

    , and .

  • 3/21/2018Kazushi SANADA

    2. Kalman filter theory

    6

    A measurement system using a Kalman filter

  • 3/21/2018Kazushi SANADA

    3. Pipeline mode; Optimized finite element modelof pipeline dynamics

    7

    pin pout

    qin qoutx

    p , q

    ρ,A, c0)(

    qp

    xpA

    tq

    f

    02

    xq

    Ac

    tp

    Equation of motion

    Continuity equation

    Interlacing grid system and linear shape function

    0 fppFqBq

    ofemofemAA

    t dd

    0 qEp ofemAc

    t

    2

    dd

    TDNU ,qq,,qq 11 q TNppp 21,p

    Finite element approximation

    State variables at grid points

  • 3/21/2018Kazushi SANADA

    3. Pipeline model; Optimized finite element modelof pipeline dynamics

    8

    The interlacing grid spacing is optimized to minimize the error function of model’s eigenvalues compared with theoretical ones.

    321

    1

    2

    3

    1

    2

    2

    1

    2

    1121 1)1(2

    1)12(

    12

    ),,(k

    i n

    ik

    i n

    ik

    i n

    iN iii

    lllJ

    Uniform Optimized

    Lc

    n 2

    ni2

    ni )12(

    ni )1(2

  • 3/21/2018Kazushi SANADA

    3. Pipeline model; Optimized finite element model of pipeline dynamics

    9

    Optimized finite element model can be used by a state space block of MATLAB/Simulink.

    GUI for automatic calculation of state space representation

  • 3/21/2018Kazushi SANADA

    4. Off-line estimation of unsteady flow rate by the Kalman filter

    10

    A test circuit of measuring unsteady flow rate

    0

    0.5

    1

    0

    0.5

    1

    0

    0.5

    1

    0

    0.5

    1

    0 50 100 150 200 2500

    0.5

    1

    f [Hz]

    |q| [

    l/min

    ]|q

    | [l/m

    in]

    |q| [

    l/min

    ]|q

    | [l/m

    in]

    |q| [

    l/min

    ]

    Drive frequency of servo valve : 100Hz

    Drive frequency of servo valve : 75hz

    Drive frequency of servo valve : 50Hz

    Drive frequency of servo valve : 25Hz

    Drive frequency of servo valve : 10Hz

    FFT analysis of unsteady flow rate estimated by the Kalman filter

    5 5.01 5.02 5.0325

    26

    27

    28

    29

    30

    t [s]

    q [l/

    min

    ]

    Kalman Filter MOC Flow Sensor

  • 3/21/2018Kazushi SANADA

    5. Real-time implementation of Kalman filter

    11

    Kalman filter is installed in a real-time computing system.

    Target pipe simulator

    Kalman filter

  • 3/21/2018Kazushi SANADA

    5. Real-time implementation of Kalman filter

    12

    21μs100μs >

  • 3/21/2018Kazushi SANADA

    5. Real-time implementation of Kalman filter

    13

    5 6 7 8 9 100

    50

    100

    SlaveMaster

    N

    Turn

    arou

    nd T

    ime

    (

    s)

    Turnaround time for various numbers of elements N(Sampling time: 100μs for N=9 or less, 1000μs for N=10)

    CPU elapsed time depends on the number of finite elements of pipeline model.

    (Kalman filter)

  • 3/21/2018Kazushi SANADA

    6. Conclusion

    • Real-time implementation of the Kalman filter is proposed to measure incompressible unsteady laminar flow rate and power of fluid flow in a pipe.

    • A steady-state Kalman filter is implemented in a real-time computing system.• It successfully works with a sampling time of 100 μs.

    Future works• to reduce computational task• to reduce the order of the pipeline model by order-reduction theory, etc.• to select suitable number of finite elements

    Acknowledgment• This work is supported by Grant-in-Aid for Scientific Research (C) (JSPS KAKENHI

    Grant Number JP17K06226).

    14

  • Thank you for your attention!

    Contact:

    3/21/2018Kazushi SANADA 15

    • Kazushi SANADA• Yokohama National University• E-mail: [email protected]

    /ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 300 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 300 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages true /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /CropMonoImages true /MonoImageMinResolution 1200 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 1200 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile () /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False

    /CreateJDFFile false /Description > /Namespace [ (Adobe) (Common) (1.0) ] /OtherNamespaces [ > /FormElements false /GenerateStructure false /IncludeBookmarks false /IncludeHyperlinks false /IncludeInteractive false /IncludeLayers false /IncludeProfiles false /MultimediaHandling /UseObjectSettings /Namespace [ (Adobe) (CreativeSuite) (2.0) ] /PDFXOutputIntentProfileSelector /DocumentCMYK /PreserveEditing true /UntaggedCMYKHandling /LeaveUntagged /UntaggedRGBHandling /UseDocumentProfile /UseDocumentBleed false >> ]>> setdistillerparams> setpagedevice