EFF IC I ENT METHOD FOR MULT I -DIMENSIONAL FUZZY REASON1 NG AND ITS APPLICATION TO FAULT DIAGNOSIS
Yoichi Hayashi, Masato Nakai · 1988
This paper proposes a fast and simple method for multi-dimensional fuzzy reasoning and presents diagnostic techniques using multi-dimensional fuzzy reasoning. The reasoning is based on the iterative use of fuzzy modus ponens and fuzzy modus tollens. The present method chooses only one multi-dimensional implication among all combinations of multi-dimensional implications so that the antecedent of multidimensional implication selected is as close to the multi-dimensional premise given as possible. The number of applications of one-dimensional implication is substantially fewer than that of Sugeno and Takagi's method. Computational complexity of the proposed method and Sugeno and Takagi's method are also considered. Furthermore, seven types of fuzzy production rules for fault diagnosis are proposed.