不完全信息下一种基于方案序列优势的多属性决策方法  

A Multi-attribute Decision-Making Method Based on Advantages of Sequential Alternatives with Incomplete Information

作　　者：潘玉厚[1] 张发明[2] 郭亚军[1] 任晓彧[1] 

机构地区：[1]东北大学工商管理学院,辽宁沈阳110004 [2]南昌大学经济管理学院,江西南昌330031

出　　处：《东北大学学报（自然科学版）》2010年第6期898-901,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：针对权重信息和属性值完全未知,而决策者仅能给出同一指标下各备选方案排序关系的特殊情况,给出了一种低成本、快捷的决策方法.提出了序列优势数的概念,并在此基础上构建出不影响排序结果的虚拟决策属性值.通过已构建的虚拟决策属性值,利用逼近理想点法(TOPSIS),构建了虚拟决策指标权重,然后利用构建的虚拟决策属性值和虚拟决策指标权重对各方案数值进行集结并给出排序.最后用一个算例证明方法的有效性.In case that the information on weight and attribute are both completely unknown and the decision-makers can only sequence all the alternatives in accordance to the same index,a rapid decision-making method is proposed at the low cost.An idea named the number of sequential advantage is presented,then the virtual attribute values are formed,which don＇t affect the sequenced results.With the virtual attribute values used,a method to approximate to ideal points（TOPSIS）is applied to framing the weight of virtual decision-making index,and the values in all alternatives are aggregated and sequenced via the virtual attribute values and the weight.A numerical example is given to verify the effectiveness of the method proposed.

关 键 词：多属性决策 虚拟属性值 虚拟指标权重 逼近理想点法 

分 类 号：C934[经济管理—管理学]