A method of self-generating fuzzy rule base via genetic algorithm

It is known that the fuzzy control rules for a control system is always built by designers with trial and error and based on their experience or some experiments. This paper introduces a genetic algorithm (GA) based method to generate a satisfactory fuzzy rule base spontaneously. With the specific structure of the chromosome, the special mutation operation and the adequate fitness function, the proposed method with GA produces a fuzzy rule base with small number of rules, suitable placement of the premise´s fuzzy sets and proper location of the consequent singletons. The generated fuzzy rule base can be the controller in a closed loop system to achieve some control objective or can be a fuzzy model to approximate an unknown nonlinear system. Finally, two examples are illustrated to show the effectiveness of the proposed method on the fuzzy control design and fuzzy modeling respectively.