Yugeng Xi is a Chair Professor of Shanghai Jiao Tong University (SJTU). He received Dr.-Ing. degree on automatic control from Technical University Munich, Germany in 1984. Since then he has been with the Department of Automation, SJTU. His research interests include predictive control theory and applications, control and optimization of large scale complex systems. He has been working in the area of predictive control for more than 35 years.?? Dewei Li is an Associate Professor of Shanghai Jiao Tong University (SJTU). He received PhD. degree on automatic control from SJTU, China in 2009. From 2011, he has been with the Department of Automation, SJTU. His research interests include predictive control theory and applications, the control of robots, intelligent systems, control and optimization of large scale complex systems. He has been working in the area of predictive control for more than 10 years.

• Preface xi1 Brief History and Basic Principles of Predictive Control 11.1 Generation and Development of Predictive Control 11.2 Basic Methodological Principles of Predictive Control 61.2.1 Prediction Model 61.2.2 Rolling Optimization 61.2.3 Feedback Correction 71.3 Contents of this Book 10References 112 Some Basic Predictive Control Algorithms 152.1 Dynamic Matrix Control (DMC) Based on the Step Response Model 152.1.1 DMC Algorithm and Implementation 152.1.2 Description of DMC in the State Space Framework 212.2 Generalized Predictive Control (GPC) Based on the Linear Difference Equation Model 252.3 Predictive Control Based on the State Space Model 322.4 Summary 37References 393 Trend Analysis and Tuning of SISO Unconstrained DMC Systems 413.1 The Internal Model Control Structure of the DMC Algorithm 413.2 Controller of DMC in the IMC Structure 483.2.1 Stability of the Controller 483.2.2 Controller with the One-Step Optimization Strategy 533.2.3 Controller for Systems with Time Delay 543.3 Filter of DMC in the IMC Structure 563.3.1 Three Feedback Correction Strategies and Corresponding Filters 563.3.2 Influence of the Filter to Robust Stability of the System 603.4 DMC Parameter Tuning Based on Trend Analysis 623.5 Summary 72References 734 Quantitative Analysis of SISO Unconstrained Predictive Control Systems 754.1 Time Domain Analysis Based on the Kleinman Controller 764.2 Coefficient Mapping of Predictive Control Systems 814.2.1 Controller of GPC in the IMC Structure 814.2.2 Minimal Form of the DMC Controller and Uniform Coefficient Mapping 864.3 Z Domain Analysis Based on Coefficient Mapping 904.3.1 Zero Coefficient Condition and the Deadbeat Property of Predictive Control Systems 904.3.2 Reduced Order Property and Stability of Predictive Control Systems 944.4 Quantitative Analysis of Predictive Control for Some Typical Systems 984.4.1 Quantitative Analysis for First-Order Systems 984.4.2 Quantitative Analysis for Second-Order Systems 1044.5 Summary 112References 1135 Predictive Control for MIMO Constrained Systems 1155.1 Unconstrained DMC for Multivariable Systems 1155.2 Constrained DMC for Multivariable Systems 1235.2.1 Formulation of the Constrained Optimization Problem in Multivariable DMC 1235.2.2 Constrained Optimization Algorithm Based on the Matrix Tearing Technique 1255.2.3 Constrained Optimization Algorithm Based on QP 1285.3 Decomposition of Online Optimization for Multivariable Predictive Control 1325.3.1 Hierarchical Predictive Control Based on Decomposition–Coordination 1335.3.2 Distributed Predictive Control 1375.3.3 Decentralized Predictive Control 1405.3.4 Comparison of Three Decomposition Algorithms 1435.4 Summary 146References 1476 Synthesis of Stable Predictive Controllers 1496.1 Fundamental Philosophy of the Qualitative Synthesis Theory of Predictive Control 1506.1.1 Relationships between MPC and Optimal Control 1506.1.2 Infinite Horizon Approximation of Online Open-Loop Finite Horizon Optimization 1526.1.3 Recursive Feasibility in Rolling Optimization 1556.1.4 Preliminary Knowledge 1576.2 Synthesis of Stable Predictive Controllers 1636.2.1 Predictive Control with Zero Terminal Constraints 1636.2.2 Predictive Control with Terminal Cost Functions 1656.2.3 Predictive Control with Terminal Set Constraints 1706.3 General Stability Conditions of Predictive Control and Suboptimality Analysis 1746.3.1 General Stability Conditions of Predictive Control 1746.3.2 Suboptimality Analysis of Predictive Control 1776.4 Summary 179References 1797 Synthesis of Robust Model Predictive Control 1817.1 Robust Predictive Control for Systems with Polytopic Uncertainties 1817.1.1 Synthesis of RMPC Based on Ellipsoidal Invariant Sets 1817.1.2 Improved RMPC with Parameter-Dependent Lyapunov Functions 1877.1.3 Synthesis of RMPC with Dual-Mode Control 1917.1.4 Synthesis of RMPC with Multistep Control Sets 1997.2 Robust Predictive Control for Systems with Disturbances 2057.2.1 Synthesis with Disturbance Invariant Sets 2057.2.2 Synthesis with Mixed H2/H∞ Performances 2097.3 Strategies for Improving Robust Predictive Controller Design 2147.3.1 Difficulties for Robust Predictive Controller Synthesis 2147.3.2 Efficient Robust Predictive Controller 2167.3.3 Off-Line Design and Online Synthesis 2207.3.4 Synthesis of the Robust Predictive Controller by QP 2237.4 Summary 227References 2288 Predictive Control for Nonlinear Systems 2318.1 General Description of Predictive Control for Nonlinear Systems 2318.2 Predictive Control for Nonlinear Systems Based on Input–Output Linearization 2358.3 Multiple Model Predictive Control Based on Fuzzy Clustering 2418.4 Neural Network Predictive Control 2488.5 Predictive Control for Hammerstein Systems 2538.6 Summary 256References 2579 Comprehensive Development of Predictive Control Algorithms and Strategies 2599.1 Predictive Control Combined with Advanced Structures 2599.1.1 Predictive Control with a Feedforward–Feedback Structure 2599.1.2 Cascade Predictive Control 2629.2 Alternative Optimization Formulation in Predictive Control 2679.2.1 Predictive Control with Infinite Norm Optimization 2679.2.2 Constrained Multiobjective Multidegree of Freedom Optimization and Satisfactory Control 2709.3 Input Parametrization of Predictive Control 2779.3.1 Blocking Strategy of Optimization Variables 2779.3.2 Predictive Functional Control 2799.4 Aggregation of the Online Optimization Variables in Predictive Control 2819.4.1 General Framework of Optimization Variable Aggregation in Predictive Control 2829.4.2 Online Optimization Variable Aggregation with Guaranteed Performances 2849.5 Summary 294References 29410 Applications of Predictive Control 29710.1 Applications of Predictive Control in Industrial Processes 29710.1.1 Industrial Application and Software Development of Predictive Control 29710.1.2 The Role of Predictive Control in Industrial Process Optimization 30010.1.3 Key Technologies of Predictive Control Implementation 30210.1.4 QDMC for a Refinery Hydrocracking Unit 30810.1.4.1 Process Description and Control System Configuration 30910.1.4.2 Problem Formulation and Variable Selection 31010.1.4.3 Plant Testing and Model Identification 31010.1.4.4 Off-Line Simulation and Design 31110.1.4.5 Online Implementation and Results 31210.2 Applications of Predictive Control in Other Fields 31310.2.1 Brief Description of Extension of Predictive Control Applications 31310.2.2 Online Optimization of a Gas Transportation Network 31810.2.2.1 Problem Description for Gas Transportation Network Optimization 31810.2.2.2 Black Box Technique and Online Optimization 32010.2.2.3 Application Example 32110.2.2.4 Hierarchical Decomposition for a Large-Scale Network 32310.2.3 Application of Predictive Control in an Automatic Train Operation System 32310.2.4 Hierarchical Predictive Control of Urban Traffic Networks 32810.2.4.1 Two-Level Hierarchical Control Framework 32810.2.4.2 Upper Level Design 32910.2.4.3 Lower Level Design 33110.2.4.4 Example and Scenarios Setting 33110.2.4.5 Results and Analysis 33210.3 Embedded Implementation of Predictive Controller with Applications 33510.3.1 QP Implementation in FPGA with Applications 33710.3.2 Neural Network QP Implementation in DSP with Applications 34310.4 Summary 347References 35111 Generalization of Predictive Control Principles 35311.1 Interpretation of Methodological Principles of Predictive Control 35311.2 Generalization of Predictive Control Principles to General Control Problems 35511.2.1 Description of Predictive Control Principles in Generalized Form 35511.2.2 Rolling Job Shop Scheduling in Flexible Manufacturing Systems 35811.2.3 Robot Rolling Path Planning in an Unknown Environment 36311.3 Summary 367References 367Index 369