Extracting Trapezoidal Membership Functions of a Fuzzy Rule System by Bacterial Algorithm

This paper presents a method of using the so-called “bacterial algorithm” [4, 5] for extracting the rules of a fuzzy system. The class of membership functions is restricted to trapezoidal, as it is general enough and widely used. The algorithm contains the bacterial mutation step allowing both the change of more than one membership function at one time, and the fine-tuning. Further operators are applied for optimising the number of rules in the base, eliminating ineffective rules and contract similar rules into a single one.