Automatic Synthesis of Rule Bases of Fuzzy Control Systems Based on Genetic Algorithms

Abstract

This chapter is dedicated to the method of the rule base (RB) automatic synthesis for the fuzzy automatic control systems (FACS) of Mamdani-type with the determination of optimal consequents based on the genetic algorithms. The obtained method allows to effectively synthesize the rule bases with optimal consequents for Mamdani-type FACSs at insufficient initial information (in conditions of a high degree of information uncertainty), at a sufficiently large number of rules when the compilation of the FACS rule base on the basis of expert knowledge is not always effective, and at the various levels of experts qualifications. The study of the efficiency of the proposed method is carried out at the development of the rule base of the fuzzy controller of Mamdani-type for the automation system of the multipurpose caterpillar mobile robot (MR) with the ability of moving on vertical and inclined ferromagnetic surfaces. The results of computer simulation confirm the high effectiveness of the designed method.