multistep model predictive control of induction machines and 3 2017-07-31¢  multistep...

Download Multistep Model Predictive Control of Induction Machines and 3 2017-07-31¢  Multistep Model Predictive

Post on 31-Mar-2020

1 views

Category:

Documents

0 download

Embed Size (px)

TRANSCRIPT

  • Lehrstuhl für Elektrische Antriebssysteme & Leistungselektronik

    Technische Universität München Professor Dr.-Ing. Ralph Kennel

    Ferdinand Grimm

    Multistep Model Predictive Control of Induction Machines and 3 Level-NPC with

    DC-Link Balancing

  • Lehrstuhl für Elektrische Antriebssysteme & Leistungselektronik

    Technische Universität München Professor Dr.-Ing. Ralph Kennel

    Arcisstraße 21, 80333 München

    Tel.: 089/289 28358 Fax: 089/289 28336 email: eat@ei.tum.de

    Ferdinand Grimm

    Multistep Model Predictive Control of Induction Machines and 3 Level-NPC with

    DC-Link Balancing

  • Multistep Model Predictive Control of Induction Machines and 3 Level-NPC with DC-Link

    Balancing

    Lehrstuhl für

    Elektrische Antriebssysteme & Leistungselektronik

    der Technischen Universität München

    Professor Dr.-Ing. Ralph Kennel

    Submitted for the Degree of Master of Science M.Sc.

    in Electrical Engineering and Information Technology (TUM)

    Ferdinand Grimm

    Born 26.04.1991 in Munich, Germany

    Frauenlobstr. 20a

    80337, Munich

    Supervision : Dr.-Ing. Zhenbin Zhang, Prof. Dr.-Ing. Fengxiang Wang

    Beginning : 15.05.2016

    End : 22.03.2017

    Date of Presentation : 31.03.2017

  • Acknowledgement

    It has been a pleasure writing my master’s thesis at the chair of electrical drive

    systems and power electronics of the Technical University of Munich.

    I would like to thank my supervisor, Dr. Zhenbin Zhang, for all the support and

    his patience while guiding me through the project and giving me valuable advice for

    the completion.

    Furthermore I would like to thank my second supervisor, Prof. Fengxiang Wang for

    giving me this wonderful opportunity to write my thesis at his institute and for all

    the help and support during my stay in China.

    I am particularly grateful to Prof. Dr. Ralph Kennel giving me the possibility to

    complete my thesis at the institute for electrical drive system and power electronics.

    My grateful thanks are to the Quanzhou institute for equipment manufacturing

    team, who helped me in many ways organize my everyday life through translations,

    valuable tips, and professional advice.

    I would also like to acknowledge the whole team of the institute for electrical drive

    systems and power electronics for providing such a great and productive working

    atmosphere.

    Ferdinand Grimm

    vii

  • Page viii

    Lehrstuhl für Elektrische Antriebssysteme & Leistungselektronik

    Technische Universität München Professor Dr.-Ing. Ralph Kennel

    Arcisstraße 21, 80333 München

    Tel.: 089/289 28358 Fax: 089/289 28336 email: eat@ei.tum.de

    MASTERTHESIS MA Nummer 03620271

    Name of Student: Ferdinand Grimm

    Frauenlobstr. 20a

    80337, Munich

    Interest of Study: Elektrotechnik & Informationstechnik

    Schwerpunkte

    Title of Thesis : Multistep Model Predictive Control of Induction

    Machines and 3 Level NPC with DC-link Balancing

    Supervision : Dr.-Ing. Zhenbin Zhang, Prof. Dr.-Ing. Fengxiang Wang

    Prof. Dr.-Ing. Ralph Kennel

  • Declaration

    The work in this thesis is based on research carried out at the Institute for Electri-

    cal Drive Systems and Power Electronics, Technische Universität München (TUM)

    supervised by Dr.-Ing Zhenbin Zhang and Prof. Dr.-Ing Fengxiang Wang. No part

    of this thesis has been submitted elsewhere for any other degree or qualification and

    it is all my own work unless referenced to the contrary in the text.

    ix

  • Contents

    Acknowledgement vii

    Declaration ix

    1 Introduction 1

    1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

    1.2 Report structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

    2 System Model 3

    2.1 Induction machine . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

    2.1.1 Space vector transformation . . . . . . . . . . . . . . . . . . . 4

    2.1.2 System equations . . . . . . . . . . . . . . . . . . . . . . . . . 8

    2.2 Inverter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

    2.2.1 3 level-neutral point clamped inverter . . . . . . . . . . . . . . 10

    2.2.2 DC-link balancing . . . . . . . . . . . . . . . . . . . . . . . . . 12

    3 Predictive Current Control 15

    3.1 Single step predictive current control . . . . . . . . . . . . . . . . . . 16

    3.1.1 Predictive current control framework . . . . . . . . . . . . . . 17

    3.1.2 Time delay compensation . . . . . . . . . . . . . . . . . . . . 22

    3.1.3 Cost function . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

    3.2 Multi step prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

    3.2.1 Formulation of the optimization problem . . . . . . . . . . . . 24

    3.2.2 Solution to the optimization problem . . . . . . . . . . . . . . 26 xi

  • Page xii CONTENTS

    4 DC-Link Balancing 33

    4.1 Multistep DC-link framework . . . . . . . . . . . . . . . . . . . . . . 34

    4.2 DC-link balancing using a priori switching state selection . . . . . . . 37

    4.3 DC-link balancing by linearization . . . . . . . . . . . . . . . . . . . . 42

    4.3.1 Linearization theory . . . . . . . . . . . . . . . . . . . . . . . 42

    4.3.2 Linearization point selection . . . . . . . . . . . . . . . . . . . 43

    4.3.3 Gradient computation . . . . . . . . . . . . . . . . . . . . . . 46

    4.3.4 Solution to the optimization problem . . . . . . . . . . . . . . 51

    4.4 Nonlinear sphere decoding . . . . . . . . . . . . . . . . . . . . . . . . 53

    4.4.1 The nonlinear sphere decoding algorithm . . . . . . . . . . . . 55

    4.4.2 Application: algorithm for control performance . . . . . . . . 58

    4.4.3 Application: algorithm for control performance and input . . 62

    4.4.4 Application: algorithm for DC-link balancing . . . . . . . . . 63

    5 Comparative Analysis 67

    5.1 Simulation setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

    5.1.1 Simulation scenario . . . . . . . . . . . . . . . . . . . . . . . . 68

    5.1.2 Figures of merit . . . . . . . . . . . . . . . . . . . . . . . . . . 69

    5.2 Simulation results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

    5.2.1 Standard sphere decoder . . . . . . . . . . . . . . . . . . . . . 71

    5.2.2 DC-link balancing using a priori selection . . . . . . . . . . . . 75

    5.2.3 DC-link balancing using linearized cost function . . . . . . . . 80

    5.2.4 DC-link balancing using nonlinear sphere decoder . . . . . . . 85

    6 Conclusion 89

    Bibliography 91

  • Page xiv CONTENTS

  • List of Figures

    2.1 Relationship between abc- and αβ coordinate frame . . . . . . . . . . 5

    2.2 Relationship between αβ- and dq coordinate frame . . . . . . . . . . 7

    2.3 Circuit diagram of the 3L-NPC converter . . . . . . . . . . . . . . . . 11

    2.4 Applicable switching states of a 3-phase 3-level inverter . . . . . . . . 12

    3.1 Block diagram of predicted current control . . . . . . . . . . . . . . . 17

    3.2 The time delay compensation problem . . . . . . . . . . . . . . . . . 22

    3.3 Solution to the time delay compensation problem . . . . . . . . . . . 22

    4.1 Grouping of the applicable switching states . . . . . . . . . . . . . . . 38

    4.2 Multistep prediction using a priori selection . . . . . . . . . . . . . . 41

    4.3 Initial value selection during steady state . . . . . . . . . . . . . . . . 45

    4.4 Initial value during acceleration . . . . . . . . . . . . . . . . . . . . . 46

    5.1 Steady state THD of the standard sphere decoder . . . . . . . . . . . 71

    5.2 Steady state performance of the standard sphere decoder . . . . . . . 72

    5.3 Performance of the standard sphere decoder . . . . . . . . . . . . . . 73

    5.4 Performance of a priori selection . . . . . . . . . . . . . . . . . . . . . 76

    5.5 Steady state THD of the a priori selection . . . . . . . . . . . . . . . 77

    5.6 Steady state performance of the a priori selection . . . . . . . . . . . 78

    5.7 Steady state THD of the linearized cost function . . . . . . . . . . . . 80

    5.8 Steady state performance of the linearized cost function . . . . . . . . 81

    5.9 Performance of the linearized cost function . . . . . . . . . . . . . . . 82

    5.10 Steady state performance of the nonlinear cost function . . . . . . . . 85

    5.11 Performance of the nonlinear cost function . . . . . . . . . . . . . . . 86 xv

  • Page xvi LIST OF FIGURES

    5.12 Steady state THD of the nonlinear cost function . . . . . . . . . . . . 88

  • Page xviii LIST OF FIGURES

  • List of Tables

    2.1 Time varying parameters of the induction machine . . . . . . . . . . 9

    2.2 Constant parameters of the induction machine . . . . . . . . . . . . . 9

    4.1 Nonlinear sphere decoder parameters for control performance . . . . . 60

    4.2 Nonlinear sphere decoder parameters for control effort . . . . . . . . . 62

    4.3 Nonlinear sphere decoder parameters for DC-link balancing . . . . . . 64

    5.1 Parameters of the simulation workstation

Recommended

View more >