摘要
针对异步电机的构造与转子故障特征,基于统计学习理论提出了信号处理技术与支持向量机故障诊断相结合的方法,以鼠笼式异步电动机为研究对象,建立了电机转子故障实验系统,并采集了电机故障信号,并使用最小二乘支持向量机(LS-SVM)进行故障分类;其次针对鼠笼式电机转子多故障分类问题,提出了快速Fourier变换、小波包分析两种不同故障信号预处理方法,将采集的定子电流信号、电机机壳振动信号分别进行分析,提取了故障特征向量,并结合SVM分类方法,实现了电机转子的故障诊断;最后,实验结果表明:基于定子电流频谱的快速Fourier变换与SVM相结合分析方法的准确判断率为93.75%,而基于db3小波分析与SVM结合分析方法的准确判断率为100%,说明了小波分析与SVM结合优越性。
Based on the structure of asynchronous motor and the characteristics of asynchronous motor faults, the combination of signal --processing technology and fault diagnosis with support vector machine （SVM） has been proposed. With the study of squirrel--cage motor fault, the rotor--fault--experiment system has been set up and the fault signals have been collected. The fault classification has been realized with the use of Least Square Support Vector Machine （LS SVM）. Then two different ways of signal-- processing technique： FFT and wavelet packet transform have been used. The collected stator current signals and motor vibration signals have been analyzed respectively and the fault--characteristic--vectors have been collected. Then, with the use of SVM classification technology, the fault diagnosis of rotor has been realized. Finally, it＇s indicated by the experimental results that the accuracy of the FFT--SVM way is 93.75%, but the accuracy of the wavelet--SVM way is 100%. So that the wavelet--SVM of classification is good.
出处
《计算机测量与控制》
北大核心
2013年第2期336-339,共4页
Computer Measurement &Control
基金
国家自然科学基金(60804022)
关键词
鼠笼电机
故障诊断
小波变换
FFT
支持向量机
squirrel--cage motor
fault diagnosis
wavelet transform
FFT
support vector machine
