Genetic algorithm combined with H∞ filtering for optimizing fuzzy rules and membership functions

This study presents a hybrid approach for determining the fuzzy rules and membership functions simultaneously. The optimization process consists of a Genetic Algorithm (GA) which determines the rule base and a H∞ Filtering method for tuning the parameters of membership functions. The procedure discussed in this study is illustrated on a simple automotive cruise control problem as a case study. By comparing nominal and optimized fuzzy controllers, we demonstrate that the hybrid algorithm, as a combination of genetic algorithm and H∞ filter, can be an effective tool for improving the performance of a fuzzy controller. In the other words, the fuzzy controller thus designed can implement simpler in the real world applications, by using a few fuzzy variables. © 2008 Asian Network for Scientific Information.