基于改进BWM的属性权重确定方法OACHSSCD

In the BWM(Best-Worst Method),the smaller the maximum deviation value ξ*,the higher the consistency of AHP.However,with the reduction of information,the ξ* calculated by the BWM becomes smaller,while the consistency of AHP becomes higher,which is contradictory to the fact that missing information leads to lower consistency of AHP.This paper proposes an improved BWM.First,the best vector obtained by comparing the best attribute with the other attributes and the worst vector ob-tained by comparing the other attributes with the worst attributes are separated.Then,the missing comparison values are supple-mented according to the consistency of the reciprocal judgment matrix,respectively.Finally,the optimal integrated weight is ob-tained according to the BWM and the arithmetic mean,and on this basis,the rationality of the improved method is verified by a theorem.Furthermore,on basis of presenting the specific steps for solving the optimal weight by using the improved BWM,and through a case study and comparative analysis,it is concluded that the improved BWM is more accurate and effective than the BWM and the Bayesian BWM.