基于信息融合证据理论的往复机械故障诊断方法  

Reserach of Fault Diagnosis Based on Information Fusion D-S Evidential Theory of Reciprocate Machinery

作　　者：付强[1] 别锋锋[1] 周国强[2] 

机构地区：[1]东北石油大学机械科学与工程学院,大庆163318 [2]东北石油大学秦皇岛分院石油井架检测国家计量认证重点实验室,秦皇岛066004

出　　处：《价值工程》2012年第1期29-30,共2页Value Engineering

基　　金：黑龙江省自然科学基金资助项目No.12511007

摘　　要：以Dempster-Shafer证据理论为基础,确定出一种基于多故障特征信息融合的诊断方法,并以往复机械故障诊断为例详细介绍该诊断方法的实施过程。诊断结果表明,经过多故障特征信息的融合,诊断结论的可信度明显提高,不确定性明显减小,该技术能较好地解决往复机械故障诊断的技术问题。According to the structure and character of reciprocating machinery,proposed a fault diagnosis method of multiple faults feature information fusion on the basis of Dempster-Shafer evidential theory which is testified detail the method＇s concrete implementation steps in solving the fault diagnosis of reciprocating machinery.In what,take the fault diagnosis of reciprocating machinery,this paper discussed in detail the executing process of the diagnosis method.It is illustrated in the diagnosis process that the degree of confidence of diagnosis conclusion was evidently improved and the uncertainty of diagnosis conclusion was evidently decreased by the fusion of multiple faults feature information,which fully indicated that this diagnosis method was effective in solving the complex problems of reciprocating machinery fault diagnosis.

关 键 词：D-S证据理论 信息融合 往复机械 故障诊断 

分 类 号：TP306[自动化与计算机技术—计算机系统结构]