Exact aggregation and defuzzification applied to a neutron fuzzy controller

An efficient algorithm that computes an exact aggregation of fuzzy sets, representing the rule consequents fired by the inference engine of a fuzzy controller, is presented. Moreover, the crisp output of the aggregated set is obtained using analytical formulae derived for the center of gravity method. The basic idea is to determine characteristics such as overlapping, intersection points and level of activation among the activated fuzzy sets in every control cycle. When compared to techniques that use discrete universes of discourse, this exact aggregation and defuzzification method requires fewer operations and renders the controller output with greater precision. Simulation results of the algorithm applied to regulate the neutron power in a research reactor model are presented.