Optimization of fuzzy membership function parameters

In the design of a generic fuzzy logic controller, the parameters associated with fuzzy membership functions are defined in an ad hoc or arbitrary manner. This paper presents a scheme to obtain these parameters automatically via the optimization of a performance measure. Numerical optimization techniques are used to obtain the optimal values of fuzzy membership function parameters such that the performance measure is minimized. An example of the design procedure is provided by applying the developed scheme to the problem of autonomous vehicle following