Comparison of Direct and Incremental Genetic Algorithm for Optimization of Ordinal Fuzzy Controllers

Conventional fuzzy logic controller is applicable when there are only two fuzzy inputs with usually one output. Complexity increases when there are more than one inputs and outputs making the system unrealizable. The ordinal structure model of fuzzy reasoning has an advantage of an easier approach of setting the rules with multiple inputs and outputs. This is achieved by giving an associated weightto each rule in the defuzzification process. An ordinal fuzzy logic controller has been designed with application for obstacle avoidance of Khepera mobile robot. Implementation show that ordinal structure fuzzy is easier to design compared to conventional fuzzy controller. However finding the best weight for each rule is a large and complex search problem. A specially tailored Genetic Algorithm (GA) approach has been proposed to find the best weight value foreach rule in the ordinal structure fuzzy controller. In this work, the comparison of direct and incremental GA for optimization of the controller is presented. Simulation results demonstrated significantly improved obstacle avoidance performance of incremental GA optimization of ordinal fuzzy controllers compared to direct GA optimized controller.