基于全局分配和局部修正策略的冲突证据融合方法

Conflict evidence fusion based on the global distribution and partial remediation strategies

提出了一种有效处理冲突证据的融合方法.该方法综合考虑了D-S证据组合规则的不足及修正数据源方法的可靠性高的特点,首先引入证据距离对冲突证据进行检验;然后计算冲突证据在证据集的比重,确定采用相关策略;最后根据不同策略对冲突证据进行融合.实验结果表明,该方法与Dempster、Yager和其他典型算法相比,在可靠性和收敛性的综合考虑上更具有优势.

In order to solve the problem of the conflicting of evidence combination, the authors propose an effective method for dealing with conflicting evidence fusion. Considering imprecise and conflicting of the experts. The method combines the advantages of global conflict allocation strategy and tactics of local conflict correction to take full account of the accuracy of the collision itself. First, introducing evidence distance to test evidence conflict, and then by obtaining the proportion of conflict evident in the evidence collection, it determines the appropriate strategy, based on different strategies for DS finally improved. The numerical study is shown that, compared with other methods, the new algorithm has more benefits in terms of reliability and convergence.