基于ENEMD与Teager能量算子的轴承早期微弱故障特征提取研究

Research on early weak fault feature extraction of rolling bearing based on ENEMD and teager energy operator

针对经验模态分解(EMD)和集合经验模态分解(EEMD)方法存在模态混叠、噪声残留以及对滚动轴承早期微弱故障特征提取效果不理想的问题,提出一种基于集成噪声重构经验模态分解(ensemble noise-reconstructed empirical mode decomposition, ENEMD)与Teager能量算子的轴承早期微弱故障特征提取方法。首先,使用ENEMD对采集信号进行分解,基于固有噪声分量处理方法实现信号降噪;其次,基于峭度与相关系数的联合准则,提取ENEMD分解的峭度值及相关系数较大的imf分量进行重构;再次,利用Teager能量算子对重构信号进行分析,从而提取到滚动轴承的早期微弱故障特征;最后,基于滚动轴承内、外圈的故障振动信号,开展所提方法与基于EMD和Teager能量算子(EMD-Teager)及基于EEMD和Teager能量算子(EEMD-Teager)方法对比实验。实验结果表明,本方法能有效提取滚动轴承早期微弱故障特征,并取得略优于EMD-Teager和EEMD-Teager能量算子的效果。

Aiming at the problems of mode aliasing,residual noise and unsatisfactory effect for extracting early weak fault features of rolling bearings with empirical mode decomposition(EMD)and ensemble empirical mode decomposition(EEMD)methods,a feature extraction method for early weak faults in rolling bearing is proposed,combined ensemble noise-reconstructed empirical mode decomposition(ENEMD)with Teager energy operator in this paper.First of all,the collection signal is decomposed with ENEMD to realize signal denoising based on the intrinsic noise component processing method.Then,the value of ENEMD decomposition and the IMF component with large correlation coefficient are extracted for reconstruction based on the joint criterion of kurtosis and correlation coefficient.After that,Teager energy operator is used to analyze the reconstructed signal and the early weak fault characteristics of rolling bearing can be extracted.Finally,the proposed method is compared with the EMD-Teager method and EEMD-Teager method based on the fault vibration signals of the inner and outer rings in rolling bearings.The experimental results show that the proposed approach can extract the early weak fault characteristics of rolling bearings effectively,and achieves slightly better results than EMD-Teager and EEMD-Teager energy operators.