作　　者： ; ; ;

机构地区： 国防科学技术大学信息系统与管理学院

出　　处： 《系统工程与电子技术》 2007年第7期1082-1085,共4页

摘　　要： 针对区间型多属性决策问题,提出了一种新的决策方法。首先通过定性打分和定量优化计算的方式得到主观赋权权值向量,再基于备选方案间偏差最大化的思想计算出客观赋权权值向量,之后将这两个向量依据决策者的需求进行组合优化,得到最终合理的权值向量。最后,利用改进TOPSIS法来对该决策问题进行优劣排序计算,在有效避免了逆序问题产生的同时增大了备选方案间的排序区分度。实验结果验证了该新方法的合理性和优越性。 According to interval multi-attribute decision-making problem, a new decision making method is presented. Firstly, a subject-weighting vector can be obtained by qualitive scoring and quantitive optimum calculation. Meanwhile, based on the thought of maximizing deviation among candidate options, a objective-weigh- ting vector can be calculated. Then, through optimized combination of the requirements of these two vector determinants, the final reasonable vector can be reached. Lastly, the decision-making problem is ranked by using the improved TOPSIS method. The rank-reversal problem can be effectively avoided while increasing differentiability of rankings among candidate options. Experimental result testifies the feasibility and advantages of this new method.

关 键 词： 多属性决策 权值向量 区间 改进 法

领　　域： [自然科学总论]