ファジィ指数型AHPモデルにおけるパラメータの最適化

In this paper, we propose to introduce the concept of exponential fuzzy number into a simple AHP(Analytical Hierarchy Process) model. The variety of the bottom “a” in the exponential fuzzy number has a big influence on the decision making of AHP method. We will discuss the optimal bottom“ a” which maximizes the difference of weights between two alternatives. In addition to the exponential fuzzy number, we let the AHP model include the concept of α-Cut. The optimal “α” means that membership functions of two alternatives do not intersect at level α. We have carried out all simulations. As a result, it was found that the tracks of intersections between two membership functions have three patterns. It is convinced that the optimal“ α” will become effective for all AHP models with exponential fuzzy number.