Intelligent design of fuzzy logic controller using NSGA-II

This work proposes a new design method of fuzzy controllers with well-formed membership functions, minimal number of fuzzy rules and optimal values of scale factors. To achieve this goal, we use multiobjective genetic algorithms that optimize simultaneously the number of fuzzy rules through the control gens, and the parameters related to membership functions, conclusions of fuzzy rules and scale factors through parametric genes. The proposed method is applied to a bio-process control problem.