A New Optimization Algorithm for Fuzzy Set Design

This paper provides an overview on the Artificial Fish Swarm Algorithm (AFSA) for the automated design and optimization of fuzzy logic controller. A new optimization method for fuzzy logic controller design is proposed. The membership functions of input and output variables are defined by six parameters, which are adjusted to maximize the performance of the controller by using AFSA. This method can improve the capability of search and convergence of algorithm. Simulation experiment on water level controller is discussed by using above method. The simulation results show that fuzzy logic controller based on AFSA avoids premature effectively and prove its feasibility.

I. Introduction

Fuzzy sets and fuzzy logic proposed by Zadeh [1] have been successfully applied in many fields to deal with uncertain and ill-defined problems. Fuzzy Logic controller (FLC) is usually designed by formulating implicit knowledge of an expert into a set of linguistic variables and fuzzy rules. However, for complex control task, obtaining the full fuzzy knowledge from an expert is often quit difficult and tedious. AFSA as a new optimization algorithm becomes a very hot topic; it offers new ideas to solve the optimization problem in signal processing [2]. It combines the concepts of clone selection and immune system to produce optimal solutions for an arbitrary problem [3]. It requires no particular knowledge about the problem structure other than the objective function itself. The design aim of the control system is to maximize some performance of the closed loop process.