Expert Aided Control System Design

Successful multivariable control system design demands knowledge, skill and creativity of the designer. Artificial intelligence can facilitate the design process by capturing much of the knowledge and some of the skill of the designer into an intelligent design tool, leaving the designer free to concentrate more on the creativity aspect of the design. This publication investigates the contribution which artificial intelligence can make to multivariable control system design. It covers all the research, design, development and testing aspects of creating the expert system. The approach is a critical one, reporting on the success as well as the shortcomings of expert system technology. Full documentation of the design software applications relevant to new and experienced users is given.

1 Artificial Intelligence and Control System Design.- 1.1 Introduction.- 1.2 Why Use an Expert System?.- 1.3 Applications of Expert Systems in CACSD.- 1.4 Description of the Design System.- 1.5 Notation.- 2 The CACSD Method.- 2.1 Introduction.- 2.2 Notation.- 2.3 Summary of Factorization Theory.- 2.4 Computing the Coprime Matrix Fractions.- 2.5 Diagonal Factorization.- 2.6 The Diagonal Factorization Algorithm.- 2.7 The QSTEP Parameter.- 2.8 The Closed Loop Poles.- 2.9 The S Domain.- 2.10 Summary.- 3 Implementation of the CACSD Package.- 3.1 Introduction.- 3.2 Generating the Quadratic Programming Problem.- 3.3 Representing the Linear Constraints.- 3.4 Solving the Quadratic Programming Problem.- 3.5 The QPSOL Algorithm.- 3.6 Ancillary Functions of the CACSD Package.- 3.7 Summary.- 4 The Expert System.- 4.1 Introduction.- 4.2 The User Interface and Philosophy.- 4.3 Explaining the Design Language and Methodology.- 4.4 Presentation of the Design Status.- 4.5 Formulating the Design Specifications.- 4.6 Expanding the Scope of the CACSD Package.- 4.7 Optimizing the Use of the CACSD Subroutines.- 4.8 Summary.- 5 Implementation of the Expert System.- 5.1 Introduction.- 5.2 Selection of the Expert System Shell.- 5.3 Communication with External Programs.- 5.4 Database Facilities.- 5.5 Structure of the Expert System.- 5.6 Summary.- 6 Sample Design Sessions.- 6.1 Introduction.- 6.2 Example 1: A Mine Milling Plant.- 6.3 Example 2: A Gyroscope.- 6.4 Other Examples.- 6.5 Summary.- 7 Use of the Expert System.- 7.1 Introduction.- 7.2 The Undergraduate Control System Design Project.- 7.3 The Postgraduate Control System Design Project.- 7.4 Summary.- 8 Implementing the Controller.- 8.1 Introduction.- 8.2 Full Controller Implementation.- 8.3 Reduced Order Controller Estimation.- 8.4 Examplesof Controller Implementation.- 8.5 Summary.- 9 Conclusions.- Appendix A MV-CXS Specifications.- A.l Requirements for the Computer.- A.2 Requirements for the Plant.- A.3 The MV-CXS Command Language Specification.- A.3.1 Commands to Select a Response.- A.3.2 Graphics Commands.- A.3.5 Miscellaneous Commands.- Appendix B The CACSD Package Interface.- Appendix C The MV-CXS Student Design Project Instructions.- References.