Generating and Tuning the Fuzzy Logic Systems Developed in Transportation Applications

Abstract

A successful application of fuzzy logic implies prior determination of shapes of membership functions of input and output variables as well as generation of a fuzzy rule base. In some applications, the final set of fuzzy rules and the choice of membership functions are defined by trial and error. Mendel (1995) claims: “Prior to 1992, all fuzzy logic systems reported in the open literature fixed the parameters of the membership functions somewhat arbitrarily, e.g., the locations and spreads of the membership functions were chosen by the designer independent of the numerical training data. Then, at the first IEEE Conference on Fuzzy Systems, held in San Diego, three different groups of researchers presented the same idea: tune the parameters of a fuzzy logic system using the numerical training data.”