Analysis and design of inference mechanisms for fuzzy feedback control

The influence of the different combinations of implications and the defuzzifications on the performance of fuzzy feedback control are examined. The characteristics of two popular implications (the min and the prod implications) and two widely used defuzzification algorithms (COA and MAX algorithms) are analyzed in terms of recoverability, change in fuzziness and reasoning strategy. With the aid of simulation of examples of fuzzy force control in milling. It is shown that not every combination of implication should be complemented only by a defuzzification such that the resultant interference mechanism does not increase the fuzziness substantially. A detailed comparitive analysis of the fuzzy milling force control approaches based on the min and prod implications and the COA and MAX defuzzifications are given