摘要
采用小波包分析和支持向量机来诊断电机故障。针对电机中常见的故障,如电机振动故障,电机转子断条故障,电机转子偏心故障等,进行频谱分析,提取故障信号在动态条件下各频带能量作为故障特征向量。构建多个最小二乘支持向量机组成的多值故障分类器,将故障特征向量作为学习样本,并且输入支持向量机进行训练,分类器可以建立故障特征向量和故障类型的映射关系,从而达到电机故障诊断的目的。
A fault diagnosis method was proposed for motor rotor broken based on wavelet packet analysis（WPA） and support vector machine（SVM）.For motor failure which often occurs,such as motor vibration failure,broken rotor bar fault,rotor eccentricity fault,and so on.After spectrum analysis,extract frequency band energy in the fault signal under dynamic conditions as the fault feature vector.A multi-classifier was developed based on LS-SVM,and the fault feature vectors as learning samples,putting into SVM and training them,the classifier could realize the mapping relationship between the fault feature vector and the fault characteristic,so as to achieve the purpose of motor fault diagnosis.
出处
《煤矿机械》
北大核心
2011年第5期241-243,共3页
Coal Mine Machinery
