中文摘要：

针对传统Dempster-Shafer证据理论合成高冲突证据时会出现融合结果不合理的问题,提出了一种新的冲突证据加权方法和融合规则。该方法引入证据向量差异度概念对证据冲突系数进行修正,并应用修正后的冲突系数计算证据集互信度矩阵和证据权重,最后对加权修正后的证据进行融合。数值算例表明,该方法可以合理有效地解决冲突证据融合的问题,相比于其他改进方法,具有更快的收敛速度和更强的鲁棒性。

英文摘要：

Traditional Dempster-Shafer theory （DST） involves counter-intuitive behaviors when combining conflict evidence. This paper proposed a new evidence weight and fusion method. It introduced evidence vector dissimilarity to modify the DST conflict coefficient based on the disadvantage analysis of DST conflict definition. Then computed new evidence set mutual support matrix and evidence weight vector using the modified DST conflict coefficient. Finally, combined all weighted evidences with DST combination rule. A numerical example demonstrates the rationality and efficiency of the proposed method in combining conflict evidence. Compared with other improved combination methods, the proposed method is robust and more efficient.