基于遗传算法和小波包分析的异步鼠笼电动机故障诊断方法探究  

Research on Fault Diagnosis Method of Asynchronous Squirrel Cage Motor Based on Genetic Algorithm and Wavelet Packet Analysis

作　　者：许允之[1] 邵昊舒 牛小玲[1] 方磊[1] 

机构地区：[1]中国矿业大学电气与动力工程学院,江苏徐州221116

出　　处：《煤矿机电》2017年第3期55-58,61,共5页Colliery Mechanical & Electrical Technology

摘　　要：在线检测电动机的故障一直是一个重点研究问题,总结近些年的研究,提出了一个通过定子的电流来检测的方法,即对监测出来的定子电流信号用小波降噪的方法进行降噪处理,得到去掉噪声后的信号,随后对处理后的信号利用小波包分析法分析,提取第四层的小波包系数进行傅里叶分析,提取其频谱中的特征量,而后对特征量进行编码。由于提取后的编码数量较大,为便于之后的判断,故对于编码后的特征向量进行遗传算法的化简,得到最简化的特征向量组,再根据最简化的特征向量组即可初步判断电动机的运行故障。该故障诊断方法的结果表明,该诊断方法可以较快地明确电动机存在的故障类型。Online detection of motor fault has been a key research problem, recent studies has been provided a current through the stator to detect method, namely for the monitoring of the stator current signal, due to the inclusion of more noise, so first with wavelet denoising method for noise reduction processing, remove the noise of signal, then the signal processing using wavelet packet analysis method, extraction on the fourth floor of the wavelet packet coefficients of Fourier analysis, extraction of characteristic quantities in the spectrum, and then the characteristic quantities encode. The extracted code number of larger, easy, after the judgment, so for coding the feature vectors were the simplification of the genetic algorithm, get the most simplified feature vector group, according to the simplified to the set of feature vectors can be preliminary judgment operation fault of a motor. The fault diagnosis method shows that the fault type can be clearly defined.

关 键 词：电动机故障诊断 小波包分析 遗传算法 特征属性 

分 类 号：TM307.1[电气工程—电机] TP206.3[自动化与计算机技术—检测技术与自动化装置]