论文部分内容阅读
针对传统D-S证据理论难以解决高度冲突证据融合问题,提出一种新的证据合成算法.将贴近度概念引入D-S证据合成中,通过证据的一致性度量计算其权重,实现冲突证据的加权融合.提出证据合成方法选择判据,将证据合成分为冲突和非冲突2类,分别采用改进算法和传统算法对证据进行融合.实例验证表明,所提出的方法信息聚焦性能优越,可以有效解决冲突证据合成问题,在解决电力系统故障诊断问题方面有良好的效果.
Aiming at the problem that traditional DS evidence theory can not solve the fusion of highly conflicting evidences, a new algorithm of evidence synthesis is proposed. The concept of closeness is introduced into the synthesis of DS evidence, and its weight is calculated through the consistency measure of evidence so as to realize the weighted fusion of conflicting evidences. Evidence synthesis method is divided into two categories: conflict and non-conflict, and the improved algorithms and traditional algorithms are used to fuse the evidence respectively.Examples show that the proposed method has superior information focusing performance and can effectively solve the conflict evidence synthesis Problem, in solving power system fault diagnosis has a good effect.