基于AdaBoost-SOM方法的电机故障诊断
|
摘要
采用一种基于AdaBoost特征选择和SOM(自组织映射)相结合的电机故障诊断方法。通过对不同电机状态的性能试验,采集驱动电机的振动信号,对信号进行时域、频域以及小波包处理,构建信号的原始特征。利用AdaBoost算法和相关性去除冗余和关联度高的特征,选取具有强区分能力的特征,采用SOM对各电机状态进行故障分类,识别电机的运行状态。试验表明,该方法能够对电机的故障特征进行有效提取,提高了对电机状态的识别能力,鲁棒性更好。
A method of motor fault diagnosis was proposed based on the combination of AdaBoost feature selection and SOM(self-organizing maps). Through the property test on the different states of motor, collecting the vibration signals of the driving motor, and processing the signals by the time domain, frequency domain, and wavelet packet, the original characteristics of the signals were constructed. Using AdaBoost algorithm and the analysis of correlation to remove the redundant and highly correlated features, selecting the features with strong discrimination capabilities and adopting SOM to classify the faults of each motor state, then the operating status of motor were identified. The experiment shows that the method can effectively extract the fault features of the motor, improve the ability to recognize the states of motor, and have a better robustness.
A method of motor fault diagnosis was proposed based on the combination of AdaBoost feature selection and SOM(self-organizing maps). Through the property test on the different states of motor, collecting the vibration signals of the driving motor, and processing the signals by the time domain, frequency domain, and wavelet packet, the original characteristics of the signals were constructed. Using AdaBoost algorithm and the analysis of correlation to remove the redundant and highly correlated features, selecting the features with strong discrimination capabilities and adopting SOM to classify the faults of each motor state, then the operating status of motor were identified. The experiment shows that the method can effectively extract the fault features of the motor, improve the ability to recognize the states of motor, and have a better robustness.
引文
[1]LIU Y,BAZZI A M.A review and comparison of fault detection and diagnosis methods for squirrel-cage induction motors:State of the art[J].ISA transactions,2017,70:400-409.
[2]王丽华,谢阳阳,张永宏,等.采用深度学习的异步电机故障诊断方法[J].西安交通大学学报,2017,51(10):128-134.
[3]李平,李学军,蒋玲莉,等.基于KPCA和PSOSVM的异步电机故障诊断[J].振动.测试与诊断,2014,34(4):616-620.
[4]吴建萍,姜斌,刘剑慰.基于小波包信息熵和小波神经网络的异步电机故障诊断[J].山东大学学报:工学版,2017,47(5):223-228.
[5]韩敏,张占奎.基于改进核主成分分析的故障检测与诊断方法[J].化工学报,2015,66(6):2139-2149.
[6]周卫庆,司风琪,徐治皋,等.基于KPCA残差方向梯度的故障检测方法及应用[J].仪器仪表学报,2017,38(10):2518-2524.
[7]王惠中,夏雨婷,乔林翰,等.关于电机故障诊断方法的优化研究[J].计算机仿真,2017,34(6):361-366.
[8]WEN X,SHAO L,FANG W,et al.Efficient feature selection and classification for vehicle detection[J].IEEE Trans.Circuits Syst.Video Techn.,2015,25(3):508-517.
[9]VAN GASSEN S,CALLEBAUT B,VAN HELDEN M J,et al.FlowSOM:Using self‐organizing maps for visualization and interpretation of cytometry data[J].Cytometry Part A,2015,87(7):636-645.
[10]杨新武,马壮,袁顺.基于弱分类器调整的多分类Adaboost算法[J].电子与信息学报,2016,38(2):373-380.
[11]刘艳,陈丽安.基于SOM的真空断路器机械故障诊断[J].电工技术学报,2017,32(5):49-54.
[12]石丽红.基于SOM算法的高维数据可视化[D].河北秦皇岛:燕山大学,2013.
[13]ULTSCH A.U*-matrix:a tool to visualize clusters in high dimensional data[J].Marburg:Fachbereich Mathematik und Informatik,2003.2518-2524.
[14]蒋伟江.基于小波包和SOM神经网络的车辆滚动轴承故障诊断[J].机械设计与研究,2012,28(6):70-73.
[15]姜高霞,王文剑.时序数据曲线排齐的相关性分析方法[J].软件学报,2014(9):2002-2017.
[2]王丽华,谢阳阳,张永宏,等.采用深度学习的异步电机故障诊断方法[J].西安交通大学学报,2017,51(10):128-134.
[3]李平,李学军,蒋玲莉,等.基于KPCA和PSOSVM的异步电机故障诊断[J].振动.测试与诊断,2014,34(4):616-620.
[4]吴建萍,姜斌,刘剑慰.基于小波包信息熵和小波神经网络的异步电机故障诊断[J].山东大学学报:工学版,2017,47(5):223-228.
[5]韩敏,张占奎.基于改进核主成分分析的故障检测与诊断方法[J].化工学报,2015,66(6):2139-2149.
[6]周卫庆,司风琪,徐治皋,等.基于KPCA残差方向梯度的故障检测方法及应用[J].仪器仪表学报,2017,38(10):2518-2524.
[7]王惠中,夏雨婷,乔林翰,等.关于电机故障诊断方法的优化研究[J].计算机仿真,2017,34(6):361-366.
[8]WEN X,SHAO L,FANG W,et al.Efficient feature selection and classification for vehicle detection[J].IEEE Trans.Circuits Syst.Video Techn.,2015,25(3):508-517.
[9]VAN GASSEN S,CALLEBAUT B,VAN HELDEN M J,et al.FlowSOM:Using self‐organizing maps for visualization and interpretation of cytometry data[J].Cytometry Part A,2015,87(7):636-645.
[10]杨新武,马壮,袁顺.基于弱分类器调整的多分类Adaboost算法[J].电子与信息学报,2016,38(2):373-380.
[11]刘艳,陈丽安.基于SOM的真空断路器机械故障诊断[J].电工技术学报,2017,32(5):49-54.
[12]石丽红.基于SOM算法的高维数据可视化[D].河北秦皇岛:燕山大学,2013.
[13]ULTSCH A.U*-matrix:a tool to visualize clusters in high dimensional data[J].Marburg:Fachbereich Mathematik und Informatik,2003.2518-2524.
[14]蒋伟江.基于小波包和SOM神经网络的车辆滚动轴承故障诊断[J].机械设计与研究,2012,28(6):70-73.
[15]姜高霞,王文剑.时序数据曲线排齐的相关性分析方法[J].软件学报,2014(9):2002-2017.