Mixed H2/H∞ Multi-Objective Control for T-S Fuzzy Systems

• Main Authors: Chung-Cheng Wang, 王忠政
• Other Authors: Jyh-Horng Chou
• Format: Others
• Language: zh-TW
• Published: 100
• Online Access: http://ndltd.ncl.edu.tw/handle/42558177775720909447

碩士 === 國立高雄應用科技大學 === 電機工程系 === 100 === A mixed H2/H∞ controller design approach is proposed for continuous time T-S fuzzy system. With multiple Lyapunov function and PDC state feedback, sufficient conditions for H2/H∞ performance are derived in terms of a lot of nonlinear inequalities. For solving those inequalities, Hybrid Taguchi Genetic Algorithm (HTGA), Non-dominated Sorting Genetic Algorithm-II (NSGA-II), and Multi- Objective Particle Swarm Optimization (MOPSO) are introduced to search the desired solutions and then the controller gains. A simple way to find desired controllers is to solve the H∞ performance inequalities first, and then H2 ones. However, for a system, its H2 and H∞ performance always mutually conflicts. They cannot achieve their best values simultaneously. So we introduce the multi-objective and Pareto front concept to solve the mixed H2/H∞ performance problem. A set of mixed H2/H∞ controllers are found for engineers. They can choose the proper one for their need in the practical cases. Finally, simulation examples are given to illustrate the feasibility of the proposed approach.