Distributed Model Predictive Control for Plant-Wide Systems

SHAOYUAN LI Shanghai Jiao Tong University, ChinaYI ZHENG Shanghai Jiao Tong University, China

• Preface xi About the Authors xvAcknowledgement xviiList of Figures xixList of Tables xxiii1 Introduction 11.1 Plant-Wide System 11.2 Control System Structure of the Plant-Wide System 31.2.1 Centralized Control 41.2.2 Decentralized Control and Hierarchical Coordinated Decentralized Control 51.2.3 Distributed Control 61.3 Predictive Control 81.3.1 What is Predictive Control 81.3.2 Advantage of Predictive Control 91.4 Distributed Predictive Control 91.4.1 Why Distributed Predictive Control 91.4.2 What is Distributed Predictive Control 101.4.3 Advantage of Distributed Predictive Control 101.4.4 Classification of DMPC 111.5 About this Book 13Part I FOUNDATION2 Model Predictive Control 192.1 Introduction 192.2 Dynamic Matrix Control 202.2.1 Step Response Model 202.2.2 Prediction 212.2.3 Optimization 222.2.4 Feedback Correction 232.2.5 DMC with Constraint 242.3 Predictive Control with the State Space Model 262.3.1 System Model 272.3.2 Performance Index 282.3.3 Prediction 282.3.4 Closed-Loop Solution 302.3.5 State Space MPC with Constraint 312.4 Dual Mode Predictive Control 332.4.1 Invariant Region 332.4.2 MPC Formulation 342.4.3 Algorithms 352.4.4 Feasibility and Stability 362.5 Conclusion 373 Control Structure of Distributed MPC 393.1 Introduction 393.2 Centralized MPC 403.3 Single-Layer Distributed MPC 413.4 Hierarchical Distributed MPC 423.5 Example of the Hierarchical DMPC Structure 433.6 Conclusion 454 Structure Model and System Decomposition 474.1 Introduction 474.2 System Mathematic Model 484.3 Structure Model and Structure Controllability 504.3.1 Structure Model 504.3.2 Function of the Structure Model in System Decomposition 514.3.3 Input–Output Accessibility 534.3.4 General Rank of the Structure Matrix 564.3.5 Structure Controllability 564.4 Related Gain Array Decomposition 584.4.1 RGA Definition 594.4.2 RGA Interpretation 604.4.3 Pairing Rules 614.5 Conclusion 63Part II UNCONSTRAINED DISTRIBUTED PREDICTIVE CONTROL5 Local Cost Optimization-based Distributed Model Predictive Control 675.1 Introduction 675.2 Local Cost Optimization-based Distributed Predictive Control 685.2.1 Problem Description 685.2.2 DMPC Formulation 695.2.3 Closed-loop Solution 725.2.4 Stability Analysis 795.2.5 Simulation Results 795.3 Distributed MPC Strategy Based on Nash Optimality 825.3.1 Formulation 835.3.2 Algorithm 865.3.3 Computational Convergence for Linear Systems 865.3.4 Nominal Stability of Distributed Model Predictive Control System 885.3.5 Performance Analysis with Single-step Horizon Control Under Communication Failure 895.3.6 Simulation Results 945.4 Conclusion 99Appendix 99Appendix A. QP problem transformation 99Appendix B. Proof of Theorem 5.1 1006 Cooperative Distributed Predictive Control 1036.1 Introduction 1036.2 Noniterative Cooperative DMPC 1046.2.1 System Description 1046.2.2 Formulation 1046.2.3 Closed-Form Solution 1076.2.4 Stability and Performance Analysis 1096.2.5 Example 1136.3 Distributed Predictive Control based on Pareto Optimality 1146.3.1 Formulation 1186.3.2 Algorithm 1196.3.3 The DMPC Algorithm Based on Plant-Wide Optimality 1196.3.4 The Convergence Analysis of the Algorithm 1216.4 Simulation 1216.5 Conclusions 1237 Networked Distributed Predictive Control with Information Structure Constraints 1257.1 Introduction 1257.2 Noniterative Networked DMPC 1267.2.1 Problem Description 1267.2.2 DMPC Formulation 1277.2.3 Closed-Form Solution 1327.2.4 Stability Analysis 1357.2.5 Analysis of Performance 1357.2.6 Numerical Validation 1377.3 Networked DMPC with Iterative Algorithm 1447.3.1 Problem Description 1447.3.2 DMPC Formulation 1457.3.3 Networked MPC Algorithm 1477.3.4 Convergence and Optimality Analysis for Networked 1507.3.5 Nominal Stability Analysis for Distributed Control Systems 1527.3.6 Simulation Study 1537.4 Conclusion 159Appendix 159Appendix A. Proof of Lemma 7.1 159Appendix B. Proof of Lemma 7.2 160Appendix C. Proof of Lemma 7.3 160Appendix D. Proof of Theorem 7.1 161Appendix E. Proof of Theorem 7.2 161Appendix F. Derivation of the QP problem (7.52) 164Part III CONSTRAINT DISTRIBUTED PREDICTIVE CONTROL8 Local Cost Optimization Based Distributed Predictive Control with Constraints 1698.1 Introduction 1698.2 Problem Description 1708.3 Stabilizing Dual Mode Noncooperative DMPC with Input Constraints 1718.3.1 Formulation 1718.3.2 Algorithm Design for Resolving Each Subsystem-based Predictive Control 1768.4 Analysis 1778.4.1 Recursive Feasibility of Each Subsystem-based Predictive Control 1778.4.2 Stability Analysis of Entire Closed-loop System 1838.5 Example 1848.5.1 The System 1848.5.2 Performance Comparison with the Centralized MPC 1858.6 Conclusion 1879 Cooperative Distributed Predictive Control with Constraints 1899.1 Introduction 1899.2 System Description 1909.3 Stabilizing Cooperative DMPC with Input Constraints 1919.3.1 Formulation 1919.3.2 Constraint C-DMPC Algorithm 1939.4 Analysis 1949.4.1 Feasibility 1949.4.2 Stability 1999.5 Simulation 2019.6 Conclusion 20810 Networked Distributed Predictive Control with Inputs and Information Structure Constraints 20910.1 Introduction 20910.2 Problem Description 21010.3 Constrained N-DMPC 21210.3.1 Formulation 21210.3.2 Algorithm Design for Resolving Each Subsystem-based Predictive Control 21810.4 Analysis 21910.4.1 Feasibility 21910.4.2 Stability 22510.5 Formulations Under Other Coordination Strategies 22710.5.1 Local Cost Optimization Based DMPC 22710.5.2 Cooperative DMPC 22810.6 Simulation Results 22910.6.1 The System 22910.6.2 Performance of Closed-loop System under the N-DMPC 23010.6.3 Performance Comparison with the Centralized MPC and the Local Cost Optimization based MPC 23110.7 Conclusions 236Part IV APPLICATION11 Hot-Rolled Strip Laminar Cooling Process with Distributed Predictive Control 23911.1 Introduction 23911.2 Laminar Cooling of Hot-rolled Strip 24011.2.1 Description 24011.2.2 Thermodynamic Model 24111.2.3 Problem Statement 24211.3 Control Strategy of HSLC 24411.3.1 State Space Model of Subsystems 24411.3.2 Design of Extended Kalman Filter 24711.3.3 Predictor 24711.3.4 Local MPC Formulation 24811.3.5 Iterative Algorithm 24911.4 Numerical Experiment 25111.4.1 Validation of Designed Model 25111.4.2 Convergence of EKF 25211.4.3 Performance of DMPC Comparing with Centralized MPC 25211.4.4 Advantages of the Proposed DMPC Framework Comparing with the Existing Method 25311.5 Experimental Results 25611.6 Conclusion 25812 High-Speed Train Control with Distributed Predictive Control 26312.1 Introduction 26312.2 System Description 26412.3 N-DMPC for High-Speed Trains 26412.3.1 Three Types of Force 26412.3.2 The Force Analysis of EMUs 26612.3.3 Model of CRH2 26712.3.4 Performance Index 27112.3.5 Optimization Problem 27212.4 Simulation Results 27212.4.1 Parameters of CRH2 27312.4.2 Simulation Matrix 27312.4.3 Results and Some Comments 27412.5 Conclusion 27813 Operation Optimization of Multitype Cooling Source System Based on DMPC 27913.1 Introduction 27913.2 Structure of Joint Cooling System 27913.3 Control Strategy of Joint Cooling System 28013.3.1 Economic Optimization Strategy 28113.3.2 Design of Distributed Model Predictive Control in Multitype Cold Source System 28313.4 Results and Analysis of Simulation 28613.5 Conclusion 292References 293Index 299