Defuzzification based on fuzzy clustering

Abstract

We develop a modified fuzzy clustering algorithm for parametric defuzzification in fuzzy rule base systems. Using examples and basic defuzzification properties we compare defuzzification by clustering with the standard defuzzification methods COG (Center of Gravity) and MOM (Mean of Maxima). Concerning fuzzy sets with forbidden zones the new method proves to be superior. We present how heuristic preprocessing and quality measures are used for appropriate parameter selection.<>