不同粒度语言评价集的融合方法研究  

Study on Fusion Method of Multi-granularity Linguistic Term Sets

作　　者：刘军[1] 郭亚军[1] 田喜龙[2] 潘德惠[1] 

机构地区：[1]东北大学工商管理学院,辽宁沈阳110004 [2]辽宁省建筑设计研究院,辽宁沈阳110005

出　　处：《东北大学学报（自然科学版）》2007年第11期1669-1672,共4页Journal of Northeastern University(Natural Science)

基　　金：国家自然科学基金资助项目(70472032)

摘　　要：研究了决策过程中将多粒度语言评价集进行融合的问题.分析了现有的两种多粒度语言评价集的融合方法,提出了一种基于隶属度函数进行的多粒度语言评价集融合新方法.该方法把语言评价用二元语义和在[0,1]区间上的有重叠的模糊数进行表示,融合过程避免了原有方法只能从粒度低的评价集向粒度高的评价集转化的不足,并且计算简单.通过具体算例证明该方法能够准确地给出不同备选方案的评价得分,有助于精确解决不同粒度语言评价集的群决策问题.The fusion problem of multi-granularity linguistic term sets for assessment in decisionmaking process is studied with two existing fusion methods discussed. A new fusion method based on membership function is proposed to express the linguistic assessment in terms of 2-tuple semantics and the overlapping fuzzy numbers over the interval [0,1 ]. Such a method thus avoids the disadvantage of original methods which can only convert the low-granularity sets into high- granularity ones, besides its simple calculation. A numerical example is given to demonstrate that the new fusion method can give the accurate score to an alternative and benefit the solutions to the problems of linguistic term sets with different granularities in group decision-making processes.

关 键 词：决策 模糊 多粒度语言评价集 隶属度 融合 

分 类 号：C934[经济管理—管理学] N945.25[自然科学总论—系统科学]