基于随机集的证据理论的改进在发电机故障诊断中的应用  

Application of Improved Evidence Theory Based on Random Set in Fault Diagnosis of Generating Motor

作　　者：徐余法[1] 李月[1,2] 陈国初[1] 王健[1,2] 

机构地区：[1]上海电机学院电气学院,上海200240 [2]华东理工大学信息科学与工程学院,上海200237

出　　处：《江南大学学报（自然科学版）》2012年第4期474-477,共4页Joural of Jiangnan University (Natural Science Edition)　

基　　金：上海市自然科学基金项目(11ZR1413900);上海市教委重点科研项目(09ZZ211);上海市教委重点学科项目(J51901);上海电机学院重点学科项目(09XKJ01)

摘　　要：在多传感器信息融合处理故障诊断问题过程中,传统证据理论对含有冲突证据的处理结果与实际相悖。文中运用随机集的方法对传统的证据理论进行改进,提出了一种新的基于证据本身的可信度权重和基于证据相似度的可信度权重的证据调整方法,并将改进后的方法应用于发电机系统的故障诊断中。结果表明,与传统证据理论相比改进后的方法更加精确地辨识出故障源,提高了诊断系统的性能。It is contrary to fact when conventional evidence theory is used to deal with conflicting evidences in multi-sensor information fusion fault diagnosis system. This paper makes an improvement to the combining rules of evidence theory based on random set theory. An evidence adjusting method using belief weights of evidence itself and belief weights of evidence similarity is proposed. And this method is used in fault diagnosis of generating motor system. The result of the experiment shows that the new method can improve the reliability and accuracy of fault diagnosis results, and enhance the performance of the system.

关 键 词：信息融合 D-S证据理论 证据相似度 随机集理论 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置]