Evolutionary fuzzy neural networks automatic design of rule based controllers of nonlinear delayed systems
暂无分享,去创建一个
An evolutionary fuzzy neural network (EFNN) is used for automatic design of rule base controllers of nonlinear delayed systems. The ideal rule base for thermal regulation is designed for a Cartesian and radial partition of the error state space. New solutions for closed-loop controller learning and the rule base adjustment of delayed systems are proposed. We show that this adjustment is an anticlockwise rotation of the rule base mapping over the error state space. The structure of a controller using EFNN for nonlinear and delayed thermal regulation converges to the advanced hypothesis.
[1] Derek A. Linkens,et al. Learning systems in intelligent control: an appraisal of fuzzy, neural and genetic algorithm control applications , 1996 .
[2] Andrew G. Barto,et al. Connectionist learning for control: an overview , 1990 .
[3] Han-Xiong Li,et al. A new methodology for designing a fuzzy logic controller , 1995, IEEE Trans. Syst. Man Cybern..