基于支持向量机的异步电机转子故障诊断  

Fault Diagnosis for Induction Motor Rotor Based on Support Vector Machine

作　　者：段阳[1] 刘松[1] 侯力[1] 张祺[1] 唐艳[1] 

机构地区：[1]四川大学制造科学与工程学院,成都610065

出　　处：《煤矿机械》2011年第3期250-252,共3页Coal Mine Machinery

摘　　要：根据异步电机发生故障时振动信号的特点,提出了一种基于小波包分解和支持向量机相结合的异步电机转子故障诊断方法。通过采用快速ICA算法对振动信号进行多通道数据融合,然后进行3层小波包分解,得到各分解节点对应频带的重构信号以及对应的能量,并将各频带的能量元素组成的特征向量作为诊断模型的特征向量,输入到LS-SVM分类器中进行故障识别和分类。诊断结果表明:采用ICA-SVM模型具有较高的分类速度和很好的故障识别率。According to the characteristics of fault vibration signals of induction motor, a fault diagnosis method was presented for motor rotor broken fault based on wavelet packet analysis and support vector machine. Through the multi-channel data fusion of vibration signals by the fast ICA algorithm and three-layer wavelet package decomposition, the reconstructed signal of each frequency ranges and the energy of each decomposed node was obtained. The eigenvector formed by energy of each band was regarded as the eigenvector of diagnosis models and was inputted into the LS-SVM classifier for fault recognition and classification. The diagnostics verify that the ICA-SVM model has a faster classification speed and high recognition rate for faults.

关 键 词：振动信号 小波包分解 支持向量机 异步电机 故障诊断 

分 类 号：TM343[电气工程—电机] TP806[自动化与计算机技术—检测技术与自动化装置]