Digital representation of fuzzy inference engine

On this paper we describe steps required to fit fuzzy control into a computer code, represented with binary numbers, by using an example with two inputs and one output. This is intended because a continuous curve for the membership function is not represented at all elements; it is discretized into m quantization levels called a-levels that depend on the number of resolution bits used. Mamdani inference is applied to a pair of inputs to obtain the weights of the inferred rules using max and min operators. We have distinguished that all of defuzziftcation methods need almostk-1 iterations according to the input spaces given by 2ⁿ where n is the number of bits used. We will introduce a new defuzzification method called Center of Slice Area Average (COSAA), on this method, we calculate the center of area of every slice that forms resultant membership function formed by an α - level and get an average from them, requiring m-1 iterations. This defuzziftcation depends on the number of discretization levels of membership functions, not on the output space, this reduces number of instructions to be executed, in consequence fewer processing time is consumed.