基于偏好矩阵的混合型多属性决策方法

针对混合型多属性决策问题,提出了一种基于偏好矩阵的决策方法。该方法从对偏好关系赋予基数值的角度思考并分析问题,通过确立标准优劣差异矩阵同时实现方案间属性优劣比较和差异规范,并充分考虑了区间型和风险型属性各自的不确定性特点;定义统一的偏好映射建立偏好矩阵,然后求解排序向量来量化方案的相对重要程度,再聚合得到总排序向量,实现方案排序。给出了应用该方法的决策原理与具体求解步骤。最后,以自动目标识别系统评估为例进行了算例分析,此算例说明了所提方法的有效性。

Method for solving hybrid multiple attribute decision making problems based on preference matrix

This paper investigates the hybrid multiple attribute decision making problems.Based on the preference matrix,a new decision making method is proposed.This new method analyzes the problems from a cardinal valued preference aspect.The standard predominance matrices are firstly built to compare each two solutions and quantify their differences,while the uncertainty of the interval and risk attributes are well considered.Using a universal mapping function,the paper establishes the preference matrix for each attribute.All local priority vectors of preference matrices are figured out to quantify the alternatives’ priority,and then be aggregated to a global priority vector which all alternatives are ranked according to.The paper explains the principle of the proposed method and then specializes each resolving step.Finally,a numerical example of automatic target recognition system evaluation is given to illustrate the new method very effetive.