不确定多指标决策问题的相对相似关系法研究

针对指标权重未知的不确定多指标决策问题,从描述区间数的相似性程度出发,提出了一种全新的区间数相对相似度和决策方案相对相似度定义,给出了区间数相对相似关系理论并得到了一些相关结果;然后在通过相对相似度对决策指标权重进行度量的基础上,采用备选决策方案在决策方案集中的总体相对相似度值大小选取最优对象并排序,以此给出了不确定多指标决策问题的相对相似关系法.通过算例分析了该算法和离差最大化算法的相异,说明了该算法的适用性和合理性.

Study on relative similarity relation method for uncertain multi-attribute decision making problem

For the uncertain multiple attribute decision-making problems with unknown of the attribute weight, the paper defines the new relative similarity degree of interval numbers and decision-making alternatives in order to describe the similarity of the interval numbers, presents the relative similarity relations theory of the interval numbers and obtains some related results. With the help of the measure to the attribute weight by the relative similarity degree of attribute values, the overall relative similarity degree value of the decision-making alternatives in the decision-making alternatives set is used to select the optimal object and ranked. Then the algorithm for the uncertain multiple attribute decision making based on relative similarity relation is presented. Finally, the similarities and differences between the algorithm we used in the paper and maximizing deviations are compared and analyzed by a numerical example, stating the applicability and rationality of the method.