A study on finding fuzzy rules for semi-active suspension controllers with genetic algorithm

This paper presents a new method to generate fuzzy controllers through genetic algorithm (GA). A GA with a new local improvement mechanism, which is based an genetic recombination in bacterial genetics, is applied to our method. This new algorithm is efficient in improvement of local portions of chromosomes. Appropriate combinations of input variables, number of fuzzy rules and parameters for membership functions are determined automatically through the GA operations. Comparisons with other conventional GAs are also discussed. To show the effectiveness of our approach, fuzzy controllers for semi-active suspension system are generated