Optimization of Takagi-Sugeno fuzzy controllers using a genetic algorithm

Genetic algorithms (GAs) have been successfully applied in several problems of fuzzy systems optimization. However, optimization of fuzzy controllers based on GAs has been restricted to simulations, mainly due the random characteristic of GAs. This paper proposes a genetic algorithm for real-time control optimization problems. The proposed GA is used to optimize the rule consequent parameters of a Takagi-Sugeno (TS) fuzzy controller. The effectiveness of the proposed optimization algorithm is tested by experiments in the position control of a driven pendulum. Experimental results show that the proposed GA can effectively tune the TS controller parameters in a real-time control problem