基于最大最小偏差与相似度的区间型群组决策方法

Interval-type Group Decision Making Method Based on Maximum and Minimum Deviation and Similarity

针对专家偏好信息为区间型的多属性群决策问题,文章提出基于最大最小偏差与区间相似度的决策方法。该方法首先根据最大最小偏差的模型求解属性权重,将专家决策矩阵集结为综合决策矩阵,其次通过构造区间数相似度计算标准区间的程度,根据各专家偏好信息贴近标准区间的程度确定专家权重,利用加权和集结各方案的综合评价值,基于相似度定义新的区间数比较方法对所有方案排序并选出最优方案。最后通过算例说明该方法的适用性和可行性。

Aiming at the problem of multi-attribute group decision making in which the expert preference information is interval,this paper proposes a decision making method based on maximum and minimum deviation and interval similarity.The method first solves the attribute weights according to the model of maximum and minimum deviation and aggregates the expert decision matrices into a comprehensive decision matrix,then calculates the standard interval by constructing the interval number similarity,and determines the expert weight according to the proximity of expert preference information to the standard interval.By weighting and aggregating the comprehensive evaluation values of each scheme,a new interval number comparison method is defined based on the similarity degree to rank all schemes and select the best scheme.Finally,an example is given to illustrate the applicability and feasibility of the proposed method.