A method for automatic rule and membership function generation by discretionary fuzzy performance function and its application to a practical system

In a practical system, the determination of fuzzy rules and membership functions depends on how skillfully the know-how of the experienced operator with an adequate comprehension of the target system can be extracted and developed into such rules and membership functions. In actuality, interviews, field research, etc. are made for collecting the know-how of the skilled operator. Particularly in determination of membership functions, it is no exaggeration to say that tuning is repeated on a trial and error basis. This paper proposes the fuzzy clustering method dependent on the manual operation data by the skilled operator and a discretionary fuzzy performance function, and an automatic generation method of rules and membership functions in case performance function is clearly known.<
>