基于EEMD和优化SVM的水电机组振动故障诊断  

Vibration Fault Diagnosis for Hydroelectric Generating Units Using the EEMD and Optimization Support Vector Machine

作　　者：秦正飞 王煜[1] 高磊[1] 汪健[1] 

机构地区：[1]昆明理工大学,云南省昆明市650092

出　　处：《水电与抽水蓄能》2016年第1期67-70,共4页Hydropower and Pumped Storage

摘　　要：针对水电机组故障具有渐变性等特征,提出了一种基于总体平均经验模态分解(EEMD)和优化支持向量机(SVM)相结合的水电机组故障智能诊断方法。利用EEMD能对机组振动信号进行自适应分解成若干本征模式分量(IMF),并能有效抑制经典经验模式分解(EMD)的端点效应以及模式混叠现象。从IMF分量中提取出来的能量特征作为输入建立优化SVM,以此来判断机组的故障状态。通过实例分析表明:建立的混合智能诊断方法的分类正确率高,能有效诊断机组存在的故障。Due to the fault attributes of Hydroelectric Generating Units are gradual change, a hybrid diagnosis model for Hydroelectric Generating Units diagnosing based on EEMD and SVM was proposed.With the EEMD method, the Hydroelectric Generating Units vibration signals are adaptive decomposed into a finite number of Intrinsic Mode（IMF）, which can effectively inhibit the end effect of classic Empirical Mode Decomposition （EMD） and the mode aliasing.The energy character vectors of every IMF component is calculated and the energy features extracted from a number of IMFs that contained the most dominant fault information are served as the input optimization of the support vector machine, then the fault state of the Hydroelectric Generating Units can be determined.An example demonstrates the high correct classification rate of the proposed method.

关 键 词：水电机组 EEMD 优化SVM 故障诊断 

分 类 号：TM312[电气工程—电机]