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。 Using wavelet packet analysis and support vector machine to diagnose motor fault. Aimed at common faults in motor, such as motor vibration fault, motor rotor broken bar fault and motor rotor eccentric fault, spectrum analysis is performed to extract energy of each band of fault signal under dynamic conditions as fault eigenvector. The multi-valued fault classifier composed of multiple least square support vector machines is constructed. The fault feature vectors are taken as learning samples and input to SVM for training. The classifier can establish the mapping relationship between fault feature vectors and fault types so as to reach the motor The purpose of troubleshooting.