基于D-S融合证据的决策新方法

针对现有基于D-S融合信息进行决策存在的问题,提出一种基于不确定融合证据的决策方法.该方法结合基元属性和非基元属性进行决策.首先给出了对候选决策集进行精简的原则,建立了焦元属性对目标决策属性的属性层面支持度以及证据层面支持度计算模型;然后对决策属性的决策向量、决策矩阵、理想状态向量进行定义,给出决策属性总体支持度的计算方法;最后对多源目标识别融合结果进行决策分析.结果证明了该方法的合理性与有效性,它同时具有对高冲突证据融合结果与低冲突证据融合结果的决策能力.

New decision making method based on D-S fusion evidence

To solve the problems in some decision making approaches, a decision making method based on uncertain fusion evidence is presented. The method combines decision making attributes of basic elements and non-basic elements to make decision. Firstly, the rules of simplifying the candidate decision making set is given, the support degrees model of attribute level and evidence level of focal elements to decision making elements are built. Then the decision making vectors, matrix and the ideal state vector are defined, the computing formula of total support degrees is given. Finally, the fusion results of multi-source information about target identification shows the validate and reasonable of the method, and it has advantages of processing the fusion result of high conflict evidence and the fusion result of low conflict evidence.