Adaptive fuzzy control: a GA approach

This paper presents practical approach in design and implementing an adaptive fuzzy control system, GA FuzzyWare (GAF), that utilizes genetic algorithm (GA) as the adaptation engine. We examine the fuzzy control side of the GAF system, which uses four-point fuzzy membership set for its efficiency on control environment. A three-phased inference engine is used for preprocess, fuzzy inference, clad postprocess. GAF also provides the capability to automatically emulate a system based on its data set. To eliminate the problem of tuning fuzzy sets and fuzzy rules that are common to many fuzzy systems, GAF uses GA to adapt the fuzzy control system. This paper, discusses how GAF applies genetic algorithm to adapt fuzzy rule based systems and the details of adaptation operators.