一种基于多粒度语言偏好矩阵的多属性群决策方法  

Approach for multi-attribute group decision-making with multi-granular linguistic preference matrices

作　　者：张园林[1] 匡兴华[1] 

机构地区：[1]国防科技大学信息系统与管理学院,长沙410073

出　　处：《控制与决策》2008年第11期1296-1300,共5页Control and Decision

基　　金：国家自然科学基金项目(70272002)

摘　　要：针对决策专家以多粒度语言偏好矩阵形式给出偏好信息的多属性群决策问题,提出一种基于二元语义一致化的多属性群决策方法.首先,构建一个基本语言偏好集作为多粒度语言一致化的参考集合;然后,采用基于二元语义的一致化处理方法将不同粒度的语言偏好信息均统一转化为相同粒度的二元语义形式,再通过二元语义的相关集结算子,对各决策专家给出的偏好信息进行集结并进行方案优选,得到满意结果;最后,通过算例说明了该方法的有效性.An approach based on two-puple linguistic to deal with multi-attribute group decision-making problem with multi-granular linguistic preference matrices is presented. In this approach, firstly, the basic linguistic term set is constructed and the computational formulae are given to uniform the multi-granular linguistic preference matrixes into the normalized representation of the two-puple linguistic. Then, based on the two-tuple order weighter averaging （T- OWA） and other aggregation operators, every individual＇s s preference information is aggregated into the group preference information and the most desirable alternative is selected. Finally, a numerical example illustrates the effectiveness of the proposed approach.

关 键 词：多粒度语言 二元语义 偏好矩阵 群决策 

分 类 号：C954[社会学—民族学]