A hybrid approach to adaptive fuzzy control based on genetic algorithms

This paper considers a hybrid approach to the design of adaptive fuzzy controllers in which two different learning algorithms are combined together to achieve an improved global design strategy. Namely, a GA is devised to optimize all the configuration parameters of the fuzzy controller, including the number Of membership functions and rules, while a Lyapunov-based adaptation law is used to perform a fast and fine tuning of the output singletons of the controller. A hardware non-linear benchmark is used to emphasize the particular effectiveness of the proposed approach in attacking experimental problems.