Generalized adaptive defuzzifier
Boubekeur Mendil, K. Benmahammed · 2002
A generalized adaptive defuzzifier (GAD) is proposed. GAD consists of two parts: (1) a rule firing engine (RFE) which fires only the rules whose firing strengths are equal to or greater than a prespecified threshold, and (2) a rule aggregation engine (RAE) which combines the fired rules with a modified version of the center of gravity (COG) defuzzifier. GAD is more general so that COG and mean of maxima (MOM) defuzzifiers can be regarded as two particular points in the parameter space of RFE. Furthermore, it makes use of the shape of consequent membership functions with low computational complexity and hardware implementation costs. The truck backing up example is used to demonstrate the effectiveness of the proposed defuzzifier.