摘要
针对复杂工况下电机轴承故障特征不明显的问题,提出了一种基于极点对称模态分解算法(Extreme-pointSymmetric Mode Decomposition,ESMD)与快速谱峭度联合分析的电机轴承故障诊断方法。首先将复杂故障信号进行ESMD分解得到若干模态分量(Intrinsic Mode Function,IMF)分量,利用信息熵与相关性选取有效IMF并由其信息熵确定信号重构的权重;利用快速峭度图自适应的确定带通滤波器的最佳滤波频带,对重构信号进行带通滤波;然后解调滤波信号分析,从平方包络谱中提取出相应故障的特征频率。最后通过试验分析表明,该方法可对故障信号进行有效降噪并提取出电机轴承故障特征,诊断出故障类型。
Aiming at the fault diagnosis of motor bearings in complex conditions,a fault diagnosis method for motor bearings based on the joint analysis of Extreme-point Symmetric Mode Decomposition( ESMD) and fast spectral kurtosis was proposed. First,a number of modal components( Intrinsic Mode Function,IMF)components were obtained by ESMD decomposition of the fault signal of motor bearing,and IMF was selected by information entropy and correlation,and the weight of the signal reconstruction was determined by its information entropy,and the best filter band of bandpass filter was selected by the fast kurtosis map,and the band pass of the reconstructed signal was carried out. Filtering,demodulation analysis,and extracting the fault characteristic frequency from the squared envelope spectrum. Finally,the experimental analysis shows that the method can effectively denoise the fault signal and extract the fault characteristics of the motor bearing,and diagnose the fault type.
引文
[1]安国庆,秦程,郭立炜,等.峭度滤波器用于电机轴承早期故障特征提取[J].电机与控制学报,2014(6):55-60.
[2]王金良,李宗军.极点对称模态分解方法:数据分析与科学探索的新途径[M].北京:高等教育出版社,2015.
[3]李伟红,田真真,龚卫国,等.改进的ESMD用于公共场所异常声音特征提取[J].仪器仪表学报,2016,37(11):2429-2437.
[4]Li H F,Wang J L,Li Z J.Application of ESMD Method to Air-Sea Flux Investigation[J].International Journal of Geosciences,2016,4(5):8-11.
[5]张淑清,徐剑涛,姜安琦,等.基于极点对称模态分解和概率神经网络的轴承故障诊断[J].中国机械工程,2017,28(4):425-431.
[6]Antoni J.Fast Computation of the Kurtogram for the Detection of Transient Faults[J].Mechanical Systems and Signal Processing,2007,21(1):108-124.
[7]姜建国,王庆.基于MEEMD和峭度-相关系数电机轴承故障诊断[J].自动化技术与应用,2018(1):65-70.
[8]蒋超,刘树林,姜锐红,等.基于快速谱峭度图的EEMD内禀模态分量选取方法[J].振动、测试与诊断,2015(6):1173-1178.
[9]朱可恒.滚动轴承振动信号特征提取及诊断方法研究[D].大连:大连理工大学,2013.
[10]Smith W A,Randall R B.Rolling Element Bearing Diagnostics U-sing the Case Western Reserve University Data:A Benchmark Study[J].Mechanical Systems&Signal Processing,2015,64-65:100-131.
