基于小波阈值降噪的轴承振动信号虚假模态剔除研究

Research on Eliminating False Mode of Bearing Vibration Signal Based on Wavelet Threshold Denoising

对于机械智能故障诊断而言,信号数据的质量至关重要。针对试验数据的优化问题,探究了时域信号中虚假模态的剔除研究。传统方法只能剔除频域中的虚假模态极点,而对时域信号中的虚假模态成分并不能相应剔除,因此提出通过小波阈值降噪除去振动信号中的噪声成分,减少由环境噪声引起的虚假模态,并通过模态参数识别和稳定图观察降噪前后虚假模态的出现情况以判断该方法效果。通过旋转机械试验台数据的分析处理,证明该方法对于剔除虚假模态具有一定的作用。

For mechanical intelligent fault diagnosis,the quality of signal data is very important.Aiming at the optimization of test data,this paper explores the elimination of false modes in time domain signals.Traditional methods can only remove the false modal poles in frequency domain,but cannot eliminate the false modal components in time domain signals.Therefore,this paper proposes to use wavelet threshold denoising to remove the noise component in vibration signal,reduce the false mode caused by environmental noise,and observe the appearance of false mode before and after noise reduction through modal parameter identification and stability diagram to judge the effect of this method.Through the analysis and processing of rotating machinery test-bed data,it is proved that this method has a certain effect on eliminating false modes.