Industrial PID Controller Tuning: With a Multiobjective Framework Using MATLAB® (Advances in Industrial Control)

Description

Industrial PID Controller Tuning presents a different view of the servo/regulator compromise that has been studied for a long time in industrial control research. Optimal tuning generally involves comparison of cost functions (e.g., a quadratic function of the error or a time-weighted absolute value of the error) but without taking advantage of available multi-objective optimization methods. The book does make use of multi-objective optimization to account for several sources of disturbance, applying them to a more realistic problem: how to select the tuning of a controller when both servo and regulator responses are important. The authors review the different deterministic multi-objective optimization methods. In order to ameliorate the consequences of the computational expense typically involved in their use―specifically the generation of multiple solutions among which the control engineer still has to choose―algorithms for two-degree-of-freedom PID control are implemented in MATLAB®. MATLAB code and a MATLAB-compatible program are provided for download and will help readers to adapt the ideas presented in the text for use in their own systems. Further practical guidance is offered by the inclusion of several examples of common industrial processes amenable to the use of the authors’ methods. Researchers interested in non-heuristic approaches to controller tuning or in decision-making after a Pareto set has been established and graduate students interested in beginning a career working with PID control and/or industrial controller tuning will find this book a valuable reference and source of ideas. Advances in Industrial Control reports and encourages the transfer of technology in control engineering. The rapid development of control technology has an impact on all areas of the control discipline. The series offers an opportunity for researchers to present an extended exposition of new work in all aspects of industrial control. Series Editor’s Foreword Acknowledgements Contents Abbreviations and Symbols Abbreviations Symbols 1 Introduction References 2 Industrial PID Control 2.1 Control System Design Scenario 2.2 Industrial Process Characteristics 2.2.1 Controlled Process Model 2.3 The PID Controller 2.3.1 Proportional Term 2.3.2 Integral Term 2.3.3 Derivative Term 2.3.4 PID Controller Formulations 2.3.5 Reference Processing and 2DoF PID 2.3.6 Conversion of 2DoF PID Controller Algorithms 2.4 Normalized Representations 2.4.1 Process Model Normalization 2.4.2 Controller Normalization References 3 PID Controller Considerations 3.1 Control System Evaluation Metrics 3.1.1 Performance 3.1.2 Robustness 3.1.3 Control Input Usage 3.2 Control System Trade-Offs 3.2.1 Servo Versus Regulation 3.2.2 Performance Versus Robustness 3.2.3 Input Versus Output Disturbances References 4 PID Controller Design 4.1 PID Controller Tuning 4.1.1 Analytical Tuning Methods 4.1.2 Tuning Based on Minimization of Performance Criteria 4.1.3 Tuning Rules for Robustness 4.2 Formalization of PID Tuning as a Multiobjective Optimization Problem 4.2.1 Cost Function and Constraint Selection 4.2.2 PID Tuning Problem Formulation for Integral Cost Functions References 5 Multiobjective Optimization 5.1 Formalization of the Multiobjective Optimization Problem 5.1.1 Definition of the Pareto Front 5.1.2 Practical Example 5.1.3 Different Approaches to Obtain the Pareto Front 5.2 Scalarization Algorithms to Find the Pareto Front 5.2.1 Weighted Sum 5.2.2 Normal Boundary Intersection 5.2.3 Normalized Normal Constraint 5.2.4 Enhanced Normalized Normal Constraint 5.3 Solution Selection from the Pareto Front 5.3.1 Visualization of the Pareto Front 5.3.2 Pareto as a Decision Model 5.3.3 Pareto Front as Part of a Decision Tool References 6 Application of the Multiobjective Approach 6.1 Comparison of the Methods to Obtain the Pareto Front 6.1.1 Performance Comparison of the Scalarization Methods 6.1.2 Analysis of the Results from the Control Theory Perspective 6.2 High-Order Benchmark Plant 6.3 LiTaO3 Thin Film Deposition Process References 7 PID Tuning as a Multiobjective Optimization Problem 7.1 Solution of the Multiobjective Optimization Tuning 7.2 Viability for Tuning Rules 7.2.1 Tuning of a PI Controller with Two Objectives 7.2.2 Tuning for a Three-Objective PID Controller 7.2.3 Comments on Creating Tuning Rules from Pareto Fronts 7.3 Database Approach for the Final Tuning 7.3.1 Example Using MOOTuning References 8 Industrial Application Examples 8.1 Continuously Stirred Tank Heater 8.1.1 Description of the Process 8.1.2 Simplified Linear Model 8.1.3 PID Control of the CSTH Considering Two Integral Cost Functions 8.1.4 PID Control of the CSTH Considering Three Integral Cost Functions 8.2 Continuously Stirred Tank Reactor 8.2.1 Description of the Process 8.2.2 Linearization 8.2.3 Controller Design 8.2.4 Validation of the Controller Designs 8.3 Final Remarks References