Study On Inconsistency Of Fuzzy Analytic Hierarchy Process

Analytic hierarchy process(AHP) is a multi-attribute decision making method based on the combination of qualitive and quatitive analysis. It has been widely applied in various fields, while it can make complex multi-attribute decision making problems simple.However, there are some shotcomings in AHP. The judgement matrix and comprehensive ranking are inconsistent, which is paid attention to by lots of scholars in relative field. FAHP is multi-attribute decision making method which is a combination of AHP and fuzzy mathematics. FAHP overcomes the problem of limited 1-9 scale in AHP to a certain extent. But the inconsistency of judgement matrix and comprhensive ranking remains. As a result, the last decison will change and be different from the previous one, that is contradictory and lead to a bad decision.Firstly we conducted a system research to the problem of the two kinds of inconsistency, as well as we studied the previous imporvement method and found effective improvement thinkings. Secondly we presented two new determination methods and one improvement method based on the least square method to solve the inconsistency problem in FAHP. Through a lot of simulation calculation, we found the example of inconsistent comprehensive ranking in FAHP and analyzed the causes of this phenomenon. At last, we put forward a improvement method based on ideal point, benchmark and ideal point’nearness degree to solve the inconsistent problem of comprehensive ranking. Then we did the empirical research on the algorithm, verified the feasibility and perfected the knowledge system of FAHP.The algorithm put forward in this paper will open up a new direction for the study of FAHP’inconsistency problem. I believe FAHP model will be more universal in practical decision making with the development of improvement thinking.