Study of Hybrid Taguchi-Particle Swarm Optimization and Its Application to Fuzzy Controller Design

Abstract

In this study a Hybrid Taguchi Particle swarm optimization with Crossover (HTPC) is proposed for designing a fuzzy controller Particle swarm optimization (PSO) easily falls into the trap of a locally optimal solution and has premature convergence shortcomings We use the Taguchi method with crossover operator to assist in PSO to avoid the premature convergence and escape the local optimum The proposed HTPC is effectively applied to solving 23 benchmark problems of global numerical optimization (which have unimodal and multimodal functions) The simulation experiments show that the proposed HTPC not only can solve problems that easily fall into the trap of locally optimal solutions during optimization for multimodal optimization problems but also can find the optimal solutions or close to the optimal solutions for unimodal optimization problems Simultaneously HTPC is compared with other algorithms The results show that HTPC has a superior performance in convergence rate than other algorithms Finally we use HTPC to tune the scaling factors of the fuzzy controller for nonlinear systems The simulation results demonstrate the validity and feasibility of the proposed methodology

Date of Award	2014 Sept 2
Original language	English
Supervisor	Tzuu-Hseng S. Li (Supervisor)