Research On Adaptive Fuzzy Logic Controllers With Variable Universe

This paper presents the studies of the variable width of the input and output universe for fuzzy controllers, its influence to the output of fuzzy logic controllers and the improvement of the final control effect through the variable universe. By analyzing the relationship between the control parameters, like scaling factors, and the factors for a variable universe [1] which is used to change the width of the universe, the equivalence of these parameters are also given. Simulation results indicate how the scaling factors affect system control through changing the width of the universe.In many situations, it is very hard to acquire the accurate universe for a fuzzy logic controller. In this paper, a scheme that uses the fuzzy inference rules to describe the law of the variation of a universe in a control process is proposed. By combining the method with neural networks, two practical fuzzy logic controllers are brought forward: one is the Adaptive Fuzzy Logic Controller with Variable Universe Based On Fuzzy Rules (AFLCVFR), the other is the Adaptive Fuzzy Logic Controller with Variable Universe BasedOn Fuzzy Neural Networks (AFLCVFNN). The nonlinear system simulation results of the two controllers and the Adaptive fuzzy controllers based on variable universe [1'2] proposed by Prof. Li Hongxing are showed. This paper also presents an improved scheme for a fuzzy controller using look-up tables of control rules [19] and its simulation results are given.