基于CEEMD的滚动轴承振动信号自适应降噪方法

针对滚动轴承故障特征信号容易被噪声掩盖难以提取的问题,提出了基于互补集合经验模态分解(CEEMD)的滚动轴承振动信号自适应降噪方法。为了准确判定噪声分量和有用信号分量的分界点,在对振动信号进行CEEMD分解后,设计了依据信噪分量自相关函数的单边波峰宽度特性自适应地判定分界点的方法。为了保证重构信号的完整性,利用改进的小波阈值降噪方法提取低频IMF分量中的高频有效信息。实验分析表明,结合改进阈值函数的CEEMD自适应降噪方法能够有效地去除故障振动信号中夹杂的噪声,并且很好地保留了滚动轴承振动信号的突变细节,达到了不错的降噪效果。

Adaptive De-noising Method of Rolling Bearing Vibration Signal Based on CEEMD

Aiming at the problem that the rolling bearing fault characteristic signal is easy to be masked by noise, an adaptive de-noising method based on complete ensemble empirical mode decomposition(CEEMD)for rolling bearing vibration signal is proposed. In order to accurately determine the boundary between the noise component and the useful signal component, after the CEEMD decomposition of the vibration signal, a method for adaptively determining the boundary point based on the single-edge peak width characteristic of the signal-to-noise component autocorrelation function is designed. In order to ensure the integrity of the reconstructed signal, the improved wavelet threshold de-noising method is used to extract the high frequency effective information in the low frequency IMF component. The experimental analysis shows that the CEEMD adaptive de-noising method combined with the improved threshold function can effectively remove the noise in the fault vibration signal, and retain the abrupt details of the rolling bearing vibration signal, achieving a good noise reduction effect.