AN OVERVIEW OF TECHNIQUES FOR GENETIC EVOLUTION OF FUZZY SYSTEMS

Genetic algorithms have recently gained notoriety as search engines of remarkable power, successfully search-ing, in reasonably short times, spaces completely intractable to most traditio-nal methods. The application of the genetic algorithm to the difficult problem of fuzzy system optimization has revealed that they are capable of refining nearly every aspect of the fuzzy system, generating near-optimal con-trollers, which exceed the capabilities of both hand designed and neurally optimized systems. A variety of research has begun to indicate that the genetic algorithm has the potential of generating highly optimized controllers without receiving any expert input whatsoever. In this paper we examine techniques used in applying the genetic algorithm to the optimization of a variety of fuzzy systems.