信息融合技术在大型发电机故障诊断中的应用  

Application of Information Fusion Technology on Fault Diagnosis of Generating Units

作　　者：王敏珍[1] 张志辉 裴玉杰[3] 

机构地区：[1]长春工程学院,吉林长春130012 [2]大唐长春第三热电厂,吉林长春130012 [3]抚顺供电公司,辽宁抚顺113000

出　　处：《计算机仿真》2012年第4期349-352,共4页Computer Simulation

基　　金：吉林省科技发展计划项目(20110422);吉林省教育厅"十二五"科技项目(吉教科合字[2011]第249号)

摘　　要：研究大型发电机故障诊断问题,大型发电机组故障具有复杂性和多样性,单从某一方面进行故障诊断,诊断结果比较低。为提高了大型发电机故障诊断准确率,提出一种基于信息融合技术的大型发电机故障诊断方法。首先对故障特征进行提取和降维,然后采用多个支持向机对大型发电机组故障进行初步诊断,获得相互独立的证据,最后对各证据采用DS证据理论融合算法进行融合,从而实现对大型发电机故障的准确诊断。仿真结果表明,采用信息融合的故障诊断系统有效地提高了大型发电机故障的诊断精度,增加故障诊断结果的置信度。Study fault diagnosis of large generator.The faults of large generating set are of complexity and diversity,the diagnosis rate is relatively low from only one aspect to carry out the fault diagnosis.In order to improve the accuracy of fault diagnosis of large generator,the paper presented a large generator fault diagnosis method based on information fusion technology.The fault feature extraction and dimensionality reduction were implemented first,and then several support vector machines were used for the initial faul diagnosis of large generator to acquire independent evidences.Then all of the evidences were fused by DS evidence theory fusion algorithm,thus realizing the large generator fault diagnosis.Simulation results show that,the information fusion fault diagnosis system can effectively improve the large generator fault diagnosis precision,and increase the confidence of fault diagnosis results.

关 键 词：支持向量机 故障诊断 信息融合 证据理论 

分 类 号：TM311[电气工程—电机]