基于CEEMDAN降噪与双谱分析的滚动轴承故障诊断

滚动轴承早期故障信号中的噪声成分会影响到故障特征的提取。为了提高含噪故障信号中滚动轴承早期故障特征提 取的准确性，将基于自适应噪声的完备经验模态分解（CEEMDAN）用于滚动轴承振动信号的降噪中，并对降噪后的轴承故障信号 进行双谱分析。结果表明：CEEMDAN可有效去除轴承振动信号中的低频噪声干扰，经CEEMDAN降噪后的不同轴承故障信号的 双谱全局图存在明显差异，根据这些差异可在宏观上对不同轴承故障加以区分；通过经CEEMDAN降噪后的不同轴承故障信号的 双谱细节图可以正确提取不同轴承故障的特征频率，从而实现对各轴承故障的有效诊断。CEEMDAN降噪结合双谱分析可为滚 动轴承故障诊断提供一种新的有效方法。

Rolling Bearing Fault Diagnosis Based on CEEMDAN De-noising and Bispectral Analysis

The noise components in the early fault signals of rolling bearings can affect the extraction of fault features. To improve the accuracy of early fault feature extraction in noisy fault signals of rolling bearings, a complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) was applied to denoise rolling bearing vibration signals, and bispectral analysis was performed on the denoised bearing fault signals. The results show that CEEMDAN can effectively remove the low-frequency noise interference in bearing vibration sig? nals. After being denoised by CEEMDAN, there are significant differences in the bispectral global maps of different bearing fault signals. These differences can be used to distinguish different bearing faults on a macro level. By using the bispectral detail maps of different bear? ing fault signals denoised by CEEMDAN, the feature frequencies of different bearing faults can be accurately extracted, thus achieving an effective diagnosis of each bearing fault. CEEMDAN denoising combined with bispectral analysis can provide a new and effective method for fault diagnosis of rolling bearings.