Rolling Bearing Fault Feature Extraction Method Based on VMD and Fast-Kurtogram

The key to fault diagnosis is to effectively extract fault features from vibration signals of transmission machinery. Based on this problem, a fault feature extraction method for rolling bearings based on VMD and fast kurtogram is proposed in this paper. The energy difference is used as the selection criterion of VMD decomposition level K. Firstly, the original signal is decomposed by VMD and select the IMF containing fault information by kurtosis criterion. Reconstruct the signal by the selected IMF. Then process the reconstructed signal with the fast kurtogram to get the reconstructed signal’s center frequency and bandwidth. Finally, the fault characteristic frequency is extracted by band-pass filtering and squared envelope spectrum. The effectiveness of this method for fault feature extraction of rolling bearings is verified by experimental data.