Data-Driven Iterative Learning Control for Discrete-Time Systems

The Springer book series Intelligent Control and Learning Systems addresses the emerging advances in intelligent control and learning systems from both mathematical theory and engineering application perspectives. It is a series of monographs and contributed volumes focusing on the in-depth exploration of learning theory in control such as iterative learning, machine learning, deep learning, and others sharing the learning concept, and their corresponding intelligent system frameworks in engineering applications. This series is featured by the comprehensive understanding and practical application of learning mechanisms. This book series involves applications in industrial engineering, control engineering, and material engineering, etc.The Intelligent Control and Learning System book series promotes the exchange of emerging theory and technology of intelligent control and learning systems between academia and industry. It aims to provide a timely reflection of the advances in intelligent control and learning systems. This book series is distinguished by the combination of the system theory and emerging topics such as machine learning, artificial intelligence, and big data. As a collection, this book series provides valuable resources to a wide audience in academia, the engineering research community, industry and anyone else looking to expand their knowledge in intelligent control and learning systems. Preface 6 Contents 8 1 Introduction 12 1.1 Learning in Control 12 1.2 Iterative Learning Control 13 1.3 PID-type ILC 14 1.4 Norm Optimization-Based ILC 16 1.5 Data-Driven Design and Analysis of ILC 19 1.6 Structure of This Monograph 22 References 22 2 Compact Form Iterative Dynamic Linearation Based DDILC 26 2.1 Introduction 26 2.2 Problem Formulation 28 2.3 Controller Design 30 2.4 Convergence Analysis 33 2.5 Simulation Study 36 2.6 Summary 39 References 39 3 DDILC with Partial Form and Full Form-Based Iterative Dynamic Linearizations 41 3.1 Introduction 41 3.2 PFIDL-Based DDILC 43 3.2.1 Problem Formulation 43 3.2.2 Controller Design 47 3.2.3 Convergence Analysis 49 3.2.4 Simulation Study 54 3.3 FFIDL-Based DDILC 56 3.3.1 Problem Formulation 56 3.3.2 Controller Design 58 3.3.3 Simulation Study 60 3.4 Summary 62 References 63 4 DDILC with State-Transition-Based Iterative Dynamic Linearization 64 4.1 Introduction 64 4.2 State-Transition-Based Iterative Dynamic Linearization 66 4.3 Lifted STIDL-Based DDILC 69 4.3.1 Controller Design 69 4.3.2 Convergence Analysis 71 4.3.3 Simulation Study 74 4.4 Nonlifted STIDL-Based DDILC 77 4.4.1 Controller Design 77 4.4.2 Convergence Analysis 79 4.4.3 Simulation Study 83 4.5 Summary 87 References 88 5 Data-Driven ILC for Systems with Package Dropouts 89 5.1 Introduction 89 5.2 Problem Formulation 90 5.3 Controller Design 92 5.4 Convergence Analysis 94 5.5 Simulation Study 100 5.6 Summary 101 References 103 6 Data-Driven ILC for Systems with Varying Trial Lengths 105 6.1 Introduction 105 6.2 Problem Formulation 107 6.3 Controller Design 107 6.4 Convergence Analysis 109 6.5 Simulation Study 114 6.6 Summary 120 References 120 7 Data-Driven ILC for Systems with Quantized Data 122 7.1 Introduction 122 7.2 Problem Formulation 123 7.3 Controller Design 124 7.4 Convergence Analysis 126 7.5 Simulation Study 134 7.6 Summary 138 References 138 8 Data-Driven ILC for Specified Point Tracking 140 8.1 Introduction 140 8.2 Problem Formulation 142 8.3 Controller Design 143 8.4 Convergence Analysis 144 8.5 Simulation Study 147 8.6 DDPTPILC Using Continuous Input Information 147 8.6.1 Problem Formulation 147 8.6.2 Controller Design 149 8.6.3 Convergence Analysis 152 8.6.4 Simulation Study 157 8.7 Summary 159 References 159 9 Higher Order Data-Driven Iterative Learning Control 161 9.1 Introduction 161 9.2 Problem Formulation 162 9.3 Controller Design 163 9.4 Convergence Analysis 164 9.5 Simulation Study 172 9.6 Summary 174 References 175 10 Constrained Data-Driven Iterative Learning Control 176 10.1 Introduction 176 10.2 Problem Formulation 178 10.3 Controller Design 179 10.4 Convergence Analysis 181 10.5 Constrained Data-Driven PTPILC 183 10.5.1 Controller Design 183 10.5.2 Convergence Analysis 185 10.6 Simulation Study 186 10.7 Summary 198 References 200 11 ESO-based Data-Driven Iterative Learning Control 202 11.1 Introduction 202 11.2 Problem Formulation 203 11.3 Controller Design 205 11.4 Convergence Analysis 207 11.5 Simulation Study 213 11.6 Summary 215 References 215 12 Event-Triggered Data-Driven Iterative Learning Control 217 12.1 Introduction 217 12.2 Problem Formulation 218 12.3 Controller Design 219 12.4 Convergence Analysis 223 12.5 Extension to MIMO Nonlinear Nonaffine Systems 224 12.5.1 Problem Formulation 224 12.5.2 Controller Design 226 12.5.3 Convergence Analysis 228 12.6 Simulation Study 230 12.7 Summary 237 References 238 This book belongs to the subject of control and systems theory. It studies a novel data-driven framework for the design and analysis of iterative learning control (ILC) for nonlinear discrete-time systems. A series of iterative dynamic linearization methods is discussed firstly to build a linear data mapping with respect of the system’s output and input between two consecutive iterations. On this basis, this work presents a series of data-driven ILC (DDILC) approaches with rigorous analysis. After that, this work also conducts significant extensions to the cases with incomplete data information, specified point tracking, higher order law, system constraint, nonrepetitive uncertainty, and event-triggered strategy to facilitate the real applications. The readers can learn the recent progress on DDILC for complex systems in practical applications. This book is intended for academic scholars, engineers, and graduate students who are interested in learning control, adaptive control, nonlinear systems, and related fields.