An Acquisition of Fuzzy Rules for Mobile Robot Using Fuzzy Classifier Systems

Abstract

We had presented an automatic generation method of fuzzy rules using hyper-cone membership functions by fuzzy classifier system (FCS), and applied to the inverted pendulum control. However, it is considered that the FCS can be applied to on-line learning of autonomous robots, etc.. In this paper, we tried that the presented FCS was applied to simulation of a mobile robot. Also, two methods which genetic operations were improved were used in order to aim at the application to the real machine.