A GA-based Sugeno-type fuzzy logic controller for the cart-pole system
Ming-Yuan Shieh, C.-W. Huang, Tzung-Shiue Li · 2002
In this paper, the genetic algorithms (GAs) are applied to find out the nearly optimal fuzzy rule-base for Sugeno-type fuzzy logic controllers (S-FLC). Basically, S-FLC is a model-based fuzzy controller with a fuzzy model identified by input-output data of an objective system. Unlike the traditional FLC, the rule number of S-FLC is drastically reduced. A GA-based S-FLC is proposed for balancing the inverted pendulum, where GAs are utilized to decide the coefficients of control rules without identifying a fuzzy model. The computer simulations depict the validity of the control method and better performance is obtained by using a genetic algorithm.