基于CEEMDAN与自适应阈值降噪的滚动轴承故障诊断

针对滚动轴承故障信息受到噪声污染而难以识别的问题，提出一种基于自适应噪声完备集合经验模态分解和自适应阈值降噪(CEEMDAN-ATD)的滚动轴承故障诊断方法。首先对原始振动信号进行CEEMDAN分解；其次利用灰色关联分析法(GRA)筛选出噪声主导和信号主导的分量；然后对噪声主导分量分别进行自适应阈值降噪(ATD)处理，并与信号主导分量进行重构；最后通过分析重构信号的Teager能量谱实现滚动轴承故障的识别。采用凯斯西储大学轴承数据对所提方法进行验证，并与完全总体经验模态分解-自适应阈值降噪(CEEMD-ATD)和CEEMDAN-小波阈值降噪(CEEMDAN-WTD)2种方法作比较，结果表明，所提方法表现出较好的自适应性和去噪效果，能够较好地服务于滚动轴承故障诊断。

Fault Diagnosis of Bearing Based on CEEMDAN and Adaptive Threshold Denoising

Aiming at the problem that the fault information of rolling bearing is polluted by noise and difficult to identify , a bearing fault diagnosis method based on adaptive noise complete ensemble empirical mode decomposition and adaptive threshold denoising( CEEMDAN-ATD ) is proposed.First of all , the original vibration signal is decomposed by CEEMDAN.Secondly , the noise-dominant components and signal-dominant components is filtered out by grey relational analysis ( GRA ) .Then , adaptive threshold denoising is performed on the noise dominant components respectively , and reconstruction is performed with the signal dominant components.Finally , the fault identification of rolling bearing is realized by analyzing the Teager energy spectrum of the reconstructed signal.The proposed method is verified by bearing data from Case Western Reserve University.Compared with the two methods of complete global empirical mode decomposition-adaptive threshold denoising( CEEMD-ATD ) and CEEMDAN-wavelet threshold denoising( CEEMDAN-WTD ), the proposed method has good adaptability and denoising effect , and can serve the rolling bearing fault diagnosis well.