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 unproved 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.
A hybrid approach to adaptive fuzzy control based on genetic algorithms / Cupertino, Francesco; Giordano, V.; Naso, David; Turchiano, Biagio. - (2004), pp. 3607-3612. (Intervento presentato al convegno 2004 IEEE International Conference on Systems, Man & Cybernetics tenutosi a The Hague, The Netherlands nel October 10-13, 2004) [10.1109/ICSMC.2004.1400902].
A hybrid approach to adaptive fuzzy control based on genetic algorithms
CUPERTINO, Francesco;NASO, David;TURCHIANO, Biagio
2004-01-01
Abstract
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 unproved 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.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
