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
|
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.
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.
引文
[1]孙娓娓,王春生,姚云飞.基于自适应遗传算法的粗糙集属性约简算法[J].计算机工程与应用,2011,47(33):49-51.
[2]余汉华,何怡刚,刘春梅.基于云-遗传算法的电动机故障诊断研究[J].电源技术,2013,37(5):808-809.
[3]杨超,王志伟.遗传算法和BP神经网络在电动机故障诊断中的应用研究[J].噪声与振动控制,2010,30(5):153-156.
[4]张亚楠,魏武,武林林.基于小波包Shannon熵SVM和遗传算法的电动机机械故障诊断[J].电力自动化设备,2010,30(1):87-91.
[5]白洁.基于遗传算法异步电动机转子故障诊断研究[J].大电机技术,2011(1):31-33.
[6]杨超,王志伟.遗传算法和BP神经网络在电动机故障诊断中的应用研究[J].噪声与振动控制,2010,30(5):154-156.
[7]吴利春.基于遗传神经网络的故障诊断算法研究[D].辽宁:辽宁大学,2012:17-20.
[8]卢胜利.开关磁阻电动机系统的故障诊断方法研究[D].徐州:中国矿业大学,2010:55-56.
[9]张晓春,许允之.基于遗传算法的电动机故障判别方法[J].煤矿机电,2014(2):73-75.
[10]张光轶,苏艳琴,孟凡娜.一种基于粗糙集理论的最优故障诊断规则获取方法[J].仪器仪表用户,2011,18(6):67-69.
[2]余汉华,何怡刚,刘春梅.基于云-遗传算法的电动机故障诊断研究[J].电源技术,2013,37(5):808-809.
[3]杨超,王志伟.遗传算法和BP神经网络在电动机故障诊断中的应用研究[J].噪声与振动控制,2010,30(5):153-156.
[4]张亚楠,魏武,武林林.基于小波包Shannon熵SVM和遗传算法的电动机机械故障诊断[J].电力自动化设备,2010,30(1):87-91.
[5]白洁.基于遗传算法异步电动机转子故障诊断研究[J].大电机技术,2011(1):31-33.
[6]杨超,王志伟.遗传算法和BP神经网络在电动机故障诊断中的应用研究[J].噪声与振动控制,2010,30(5):154-156.
[7]吴利春.基于遗传神经网络的故障诊断算法研究[D].辽宁:辽宁大学,2012:17-20.
[8]卢胜利.开关磁阻电动机系统的故障诊断方法研究[D].徐州:中国矿业大学,2010:55-56.
[9]张晓春,许允之.基于遗传算法的电动机故障判别方法[J].煤矿机电,2014(2):73-75.
[10]张光轶,苏艳琴,孟凡娜.一种基于粗糙集理论的最优故障诊断规则获取方法[J].仪器仪表用户,2011,18(6):67-69.