基于改进LS-SVM的异步电机转子故障诊断

为了提高异步电动机转子故障的诊断精度,给出了一种基于改进最小二乘支持向量机(LS-SVM)的多故障分类算法。首先运用FFT处理电机的定子电流信号得到信号频谱图,从中提取故障特征向量;然后将特征向量送入改进算法进行故障诊断时,在原有多分类算法的基础上引入层次分析法确定故障类别的权重,根据权重值确定故障的诊断顺序,依次进行故障分类。实验表明,改进算法用于故障诊断节省了诊断时间,提高了诊断精度,具有很好的推广前景。

Rotor Fault Diagnosis of Asynchronous Motor Based on Improved LS-SVM

In order to improve rotor fault diagnosis accuracy of the asynchronous motor,a multi-class classification algorithm based on improved Least Square Support Vector Machine (LS-SVM)is proposed. First the fault character vectors are collected from the signal spectrum that is acquired from the signals of the motor stator current fault by FFT. Then when the feature vectors are used as the inputs of the improved algorithm for fault diagnosis,the improved algorithm confirms the weight of all faults with analytic hierarchy process,determines the order of the fault diagnosis in accordance with the weight and achieves the fault classification in turn on the basis of the former multi-class algorithm. Experimental results show that the improved algorithm saves time and improves the diagnosis accuracy when it is used for fault diagnosis and it has a bright prospect for generalization.