CEEMD和MCKD的滚动轴承早期故障特征提取
Early Fault Feature Extraction of Rolling Bearings Applying CEEMD-MCKD

当滚动轴承处于早期故障阶段的时候，受环境噪声和信号衰减的影响，滚动轴承振动信号特征频率成分难以精确提取，并且在信噪比较低时CEEMD不能很好提取微弱故障。针对上述问题，提出了基于互补集合经验模态分解（Complementary ensemble empirical mode decomposition，CEEMD）和最大相关峭度解卷积（Maxim correlated kurtosis deconvolution，MCKD）相结合的故障特征提取方法（CEEMD-MCKD）。两种方法的结合有效解决了CEEMD分解后无法提取出淹没在背景噪声中微弱信号特征的问题，又保持了信号的完备性，避免了有用信息的损失。通过仿真和试验验证了该方法的有效性及优点。
Early fault feature of rolling bearings is very weak and affected by environmental noise and the signal attenuation, the vibration signal characteristic frequency of the rolling bearing is difficult to be accurately extracted; and the background noise is very strong, which will affect the ability of the CEEMD to extract weak signal feature. Aiming at this problem, the paper put forward an integrated diagnosis method based on the complementary ensemble empirical mode decomposition (CEEMD) and maxim correlated kurtosis deconvolution (MCKD). The combination of the two methods effectively solves the problem of weak signal characteristics that cannot be submerged in background noise after CEEMD decomposition, and the completeness of the signal is maintained, and the loss of useful information is avoided. The simulation signal and experimental data prove the effectiveness and advantages of the proposed method