Two new defuzzification techniques
Daniel Brubaker · Proceedings of WESCON '93 · 2002
The term defuzzification in a fuzzy rule-base system refers to the transforming of an action designated in fuzzy terms by the rule-base into a crisp value that can be executed. This paper discusses briefly the three currently most popular defuzzification techniques-the maximizer, centroid and singleton-with their strengths and shortcomings, and then introduces two new techniques, one quite general, and the other specific to relatively small, embedded systems.