層級分析法(Analytic Hierarchy Process, AHP)廣為應用來解決不確定性和多屬性的問題。由於模糊理論(Fuzzy Theory)可以處理主觀認知上的不確定性,Buckley在1985年時提出了模糊層級分析法(Fuzzy Analytic Hierarchy Process, FAHP)。 在實際回收模糊層級分析法(FAHP)問卷時,常會發生無法通過一致性檢定。因此我們應用基因演算法(Genetic Algorithm)來改善模糊成對比較矩陣之不可接受一致性,將矩陣轉變為一致性小於0.1。其實驗結果也將與相關文獻方法做比較,以期望此研究能作為主要使用FAHP問卷之決策者的參考。
Analytic Hierarchy Process (AHP) is widely applied to solve uncertainty and multi-attribute problem. Due to fuzzy theory can deal with the uncertainty of subjective cognition, Buckley proposed Fuzzy Analytic Hierarchy Process (FAHP) in 1985. In practice, when we retrieve FAHP questionnaires, it often result inconsistency. We employ Genetic Algorithms to improve unacceptable consistency of fuzzy pairwise comparison matrices and let its consistency can be less than 0.1. Finally, the result will compare with related researches’ methods. We expect this research can be manipulated to assist every decision maker in using FAHP questionnaires.