Learning Control of Multimodal Systems by Fuzzy Automata
暂无分享,去创建一个
The random search method is well known as an optimization technique by which a global search of multimodal systems can be executed, but its convergence characteristics are not good. In order to improve the convergence characteristics of the random search method, an idea of the modification of the search probability may be used. There is the method of learning control using stochastic automata [1] as a method based on this idea. The proposed method of learning control using fuzzy automata in which the membership function [2] is used instead of the probability is more simple in the learning algorithm and able to realize more clear self-organizing operation as compared with that using stochastic automata.
[1] King-Sun Fu,et al. A Formulation of Fuzzy Automata and Its Application as a Model of Learning Systems , 1969, IEEE Trans. Syst. Sci. Cybern..
[2] Donald Macko. Natural states and past-determinism of general time systems , 1971, Inf. Sci..
[3] Kiyoji Asai,et al. A method for optimizing control of multimodal systems using fuzzy automata , 1971, Inf. Sci..
[4] King-Sun Fu,et al. A variable structure automaton used as a multi-modal searching technique , 1965 .