基于信息融合技术的水电机组振动故障诊断  

Vibration Fault Diagnosis of Hydroelectric Generating Sets Based on Information Fusion Technology

作　　者：熊伟[1] 徐绍坤[1] 程加堂[1] 

机构地区：[1]红河学院工学院,云南蒙自661199

出　　处：《人民黄河》2014年第2期124-125,128,共3页Yellow River

摘　　要：针对传统方法在水电机组振动故障诊断中存在准确度不高的问题,引入了粒子群神经网络和证据理论相结合的故障诊断方法。对水电机组振动故障的不同征兆域,采用2个并行的粒子群神经网络进行局部诊断,以获得彼此独立的证据,再由证据理论对各证据进行融合。结果表明,该方法可有效地提高诊断可信度,减少诊断的不确定性。Abstract： For the reasons of low vibration fault diagnosis accuracy of traditional diagnosis methods on hydroelectric generating sets, a method of particle swarm optimization-neural network model combined with evidence theory was applied. The two parallel particle swarm optimization-neural networks were used to carry on local fault diagnosis and acquire independent evidences each other for the different vibration fault symptom domains of hydroelectric generating sets, then the evidence theory was employed to fuse evidences. Experimental results show that the method is good to im- prove the reliability of the diagnosis and decrease the diagnostic uncertainty.

关 键 词：信息融合 水电机组 故障诊断 证据理论 粒子群神经网络 

分 类 号：TV32[水利工程—水工结构工程]