基于EEMD能量特征识别的电动机故障诊断技术研究  

Research on Motor Fault Diagnosis Technology Based on the Energy Feature Recognition of EEMD

作　　者：王博磊 曹伟 王涛 巩振泉 WANG Bolei;CAO Wei;WANG Tao;GONG Zhenquan(HEBEIJi-Yan Energy and Technology Research Institute Co.,Ltd.;HCIG XingTai Thermal Power Co.,Ltd.)

机构地区：[1]河北冀研能源科学技术研究院有限公司,050051 [2]河北建投邢台热电有限责任公司,054400

出　　处：《电机技术》2020年第5期23-25,30,共4页Electrical Machinery Technology

摘　　要：提出了一种基于能量特征识别的电动机故障诊断方法。该方法首先应用集合经验模态分解(EngembleEmpiricalModeDecomposition,简称EEMD)得到信号的IMF分量,然后使用IMF分量的能量比值和频率段的能量比值构成特征向量,并输入到支持向量机(Support Vector Machines,简称SVM)中进行识别。通过试验证明:基于EEMD能量特征的电动机故障诊断方法能准确地识别出故障及其类型。A method of motor fault diagnosis based on the energy feature recognition is suggested. First, the ensemble empirical mode decomposition(EEMD) is applied obtain IMF component of the signal, and then, the energy ratio of the IMF component and energy ratio of the frequency band are used to form a feature vector, which is input to a support vector machine(SVM) for fault diagnosis. The results show that the motor fault diagnosis method based on the EEMD is able to identify the fault type effectively and accurately.

关 键 词：电机故障诊断 集合经验模态分解 能量特征 支持向量机 

分 类 号：TM307[电气工程—电机]