Complete design of fuzzy logic systems using genetic algorithms

The paper presents a general method for constructing accurate high-dimensional fuzzy logic systems (FLSs). Generally, the design of FLSs involves determination of the number of fuzzy rules, the structure of the rules, and membership function parameters. Most techniques treat these parts separately, which may result in a suboptimal solution. We propose to optimize all three parts simultaneously using genetic algorithm (GA) techniques. While GAs are very robust with respect to avoiding local minima, they can be slow in refining the solution once near the optimum. Thus, the FLS obtained from GA search is further fine-tuned using a conjugate gradient method. The advantages of the proposed method are demonstrated through a comparison with other fuzzy modeling techniques and feedforward neural networks on modeling a nonlinear dynamic system, and industrial process