Fuzzy control-engineering for empiricists

The author discusses the development of fuzzy sets in control. The importance of fuzzy sets in control is that they support the formulation of control rules that reflect an operator's natural ambiguity about the meaning of linguistic values, avoiding forced and specious precision in the specification of control rules. In addition, because sharp transitions across set boundaries are eliminated, then the changes in control actions can be smooth and graduated. The key elements in the design of a fuzzy controller are the definitions of fuzzy sets for the input and control variables, and the IF-THEN rules that link these variables. Given that the sets and rules have been defined, then the operation of the controller follows a basic three-step sequence: fuzzification, inference and defuzzification. The author discusses this using the example of a simple controller with two input variables. The author also discusses artificial neural nets and fuzzy control applications.<
>