A Fault Diagnosis Method of Rolling Bearing Based on CEEMDAN and KFCM

Considering that the signals generated by faulty rolling bearings are often nonstationary and nonlinear. In this paper, a hybrid method based on Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) and Kernel Fuzzy C-Means (KFCM) algorithm for rolling bearing fault diagnosis is proposed. First, the signal is decomposed to obtain intrinsic mode functions (IMFs) based on CEEMDAN method. Second, feature vector matrices of rolling bearings are established through the Singular Value Decomposition (SVD) of IMFs. Finally, we utilize KFCM algorithm and Hamming nearness degree to diagnose the type of faults. The experimental result proves that the proposed fault diagnosis approach is of high performance.