Defuzzification in Fuzzy Controllers

Article type: Research Article

Authors: Hellendoorn, Hans | Thomas, Christoph

Affiliations: Siemens AG, ZFE ST SN42, Otto-Hahn-Ring 6, W-8000 Munich 83, Germany

Abstract: An important subject in fuzzy control theory is tuning of a fuzzy controller. If one wants to tune a fuzzy controller, one can focus on the choice of rules, membership functions, number of input and output fuzzy sets and their degree of overlapping, implication, and connection operations, and defuzzification method. All these choices are closely related and in no way independent of each other. We describe six important defuzzification methods and their respective merits and shortcomings, dependent on the rules, domains, etc. Further, we give an alternative approach for the case in which the output fuzzy sets have different shapes or are asymmetrical. We illustrate this by several examples.

DOI: 10.3233/IFS-1993-1202

Journal: Journal of Intelligent and Fuzzy Systems, vol. 1, no. 2, pp. 109-123, 1993