Design of fuzzy logic controllers using genetic algorithms

Since membership functions and fuzzy control rules are interdependent in designing a fuzzy logic controller (FLC), a GA-based approach is proposed for simultaneous design of these two components. With triangular membership functions, the left and right widths of these functions, the locations of their peaks, and the output fuzzy set corresponding to every possible combination of input fuzzy sets are then chosen as parameters to be optimized. In a proportional scaling method, these parameters are then transformed into real-coded chromosomes, over which arithmetical crossover and nonuniform mutation are implemented. Meanwhile, enlarged sampling space and a ranking mechanism are also be used in the evolution process. To show the application of the proposed method, a cart-centering example is given. From the simulation results, we find that the designed FLC is robust and can drive the cart system from any given initial state to the desired final state, which verifies the feasibility and validity of the proposed method