基于幂Pignistic概率距离的加权证据组合方法

Weighted evidence combination method based on power-Pignistic probability distance

针对现有的相似性/相异性测度在量化证据冲突时存在的不足,定义一种新的被称为幂Pignistic概率距离的相异性测度,并提出基于幂Pignistic概率距离的加权证据组合方法。该方法通过幂Pignistic概率距离量化两证据之间的冲突程度,然后建立相似性矩阵并求得各证据的可信度,再用加权平均法修正证据,最后利用Dempster规则进行组合。数值算例的结果表明,所提方法是合理有效的。

To deal with the shortcomings encountered when using the existing similarity/dissimilarity measures to quan-tify evidence conflict, a new dissimilarity measure called power-Pignistic probability distance was defined. Furthermore, a weighted evidence combination method was proposed based on power-Pignistic probability distance. The conflict de-gree between two pieces of evidence was quantified by the power-Pignistic probability distance. After that, a similarity measure matrix was constructed, based on which the credibility of evidence was obtained. Then the weighted average method was used to revise the evidence. Finally, the fusion was accomplished by using Dempster’s rule. The results of the numerical examples show the efficiency and rationality of the proposed method.