模糊多属性格序决策的冗余指标的消除策略

Fuzzy Multiple Attribute Lattice Decision Making Method Based on the Elimination of Redundant Similarity Index

在模糊多属性决策中,属性权重的确定对于整个评价工作有十分重要的意义.如果评价属性数量过多,指标间的相关性将影响评价的科学性和公平性.本文建立了评价值为梯形模糊数的"相似"概念和模糊相似评价模型,并基于格序决策的理论,得到了一种新的模糊格序决策方法.结合传统的TOPSIS方法,通过计算将各方案的属性值的中心进行加权后与正负理想中心的贴近度的大小,实现备选方案的格序化排序.实例分析的结果表明:方法合理、易行.

The determination of indicator weights is very important to the entire evaluation in the fuzzy multiple attribute decision. Especially in the project evaluation, it is not evitable that there is a certain correlation due to the excessive number of indicators. Study found that some correlations among the indexes will seriously affect the scientific and fairness of evaluation. The concepts of similar and the model of fuzzy similarity evaluation are built in this paper, and based on lattice decision, a new fuzzy lattice decision making is proposed. Combining with the traditional TOPSIS method, the close degrees between the weighted centers of attribute values of every program and positive ideal and negative ideal are calculated in order to achieve the lattice order of alternative schemes. The analytical results of an illustrative example show that this approach is reasonable and feasible.