基于证据理论的多信息融合故障诊断研究  

Study on the Multi-information Fusion Fault Diagnosis Based on Evidential Theory

作　　者：李若琼[1] 李欣[1] 董海鹰[1] 

机构地区：[1]兰州交通大学自动化与电气工程学院,兰州730070

出　　处：《自动化仪表》2006年第12期1-4,共4页Process Automation Instrumentation

基　　金：甘肃省自然基金资助项目(编号:3ES041-A25-025)

摘　　要：针对列车智能控制系统故障诊断中的多故障特征信息输入时的时变、不确定性和空间分布性,从D-S证据理论的基本概念和证据的融合推理方法出发,提出了一种列车智能控制系统多信息融合故障诊断的系统结构。讨论了基于D-S证据理论的列车智能控制系统多信息融合故障诊断方法。故障诊断实例的结果表明该方法能够有效地提高诊断的可信度,减小诊断的不确定性。In fault diagnosis of train intelligent control system, aiming at the time varying, uncertainty and space distribution of muhi-fauh characteristic information input, on the basis of the basic conception and evidential fusion reasoning method of Dempster · Shafer（ D-S） evidential theory, a systematic structure for fault diagnosis of multi-information fusion in train intelligent control system is presented. The method of fault diagnosis of multi-information fusion in train intelligent control system based on D-S evidential theory is discussed. Results of examples indicate that this method effectively improves the credibility and reduces the uncertainty of diagnosis.

关 键 词：证据理论 信息融合 智能控制 故障诊断 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]