A fuzzy modeling of very large scale system using genetic algorithm and a multiple-representing method

The paper presents a new approach to fuzzy modeling of very large scale systems using a genetic algorithm. A new method for obtaining multiple phenotypes from a chromosome, i.e. a multiple-representing method, for expanding exploratory space is also proposed. The new method maintains the diversity of the phenotypes and is effective in exploring the vast solution space. A simulation of fuzzy modeling to identify a multi-input nonlinear system is done to show the effectiveness of the proposed method.<
>