METHOD AND APPARATUS FOR THE CONTROL OF FUZZY LOGIC USING NEURAL NETWORK

摘要

The fuzzy control circuit using neuro circuit network comprises: a plurality of first minimum value detecting units having a first fuzzy comparing unit for generating a first selection signal and a first selection unit for selecting the smaller one of the membership function value of the actual value and the membership function value of each rule in response to the first selection signal; a plurality of first maximum value detecting units having a second fuzzy comparing unit for generating a second selection signal, a second selection unit for selecting the greater one of the output value and the previous maximum value in response to the second selection signal and a gate for feeding back the present output value to the second fuzzy comparing unit; a latch unit for latching the output of the first maximum value detecting unit; a plurality of second minimum value detecting units having a third fuzzy comparing unit for generating a third selection signal and a third selection unit for selecting the smaller one of the latched value and the membership function value of each control quantity; a plurality of second maximum value detecting units having a fourth fuzzy comparing unit for generating a fourth selection signal and a fourth selection unit for selecting the greater one of the output value and the maximum value of previous stage; and a defuzzification unit having a dichotomizing value outputting unit for generating a dichotomizing value, an accumulating unit for generating an accumulated value and a fuzzy comparing unit.