Fault feature extraction of rolling bearing based on GWO optimized SVMD

Rolling bearings typically operate in tough and complex working environments, and the fault pulse characteristics implied in the vibration signals are frequently interfered with by random noise, making fault feature extraction difficult. To address this issue, this paper provides a fault feature extraction method based on the Grey Wolf algorithm (GWO) for optimizing Successive Variational Mode Decomposition (SVMD). This method uses the minimum fuzzy entropy as the fitness function of the GWO and employs the GWO to adaptively iteratively search for the optimal SVMD balance parameter for signal decomposition, before selecting the Intrinsic Mode Function (IMF) with the maximum kurtosis as the target IMF and performing envelope demodulation analysis on it to accurately extract fault feature information. The suggested method outperforms unoptimized SVMD and Variational Mode Decomposition (VMD) algorithms in terms of computing efficiency and can highlight fault feature components, and the experimental results validate the GWO-SVMD algorithm suggested in this paper.