Bearing fault diagnosis method based on EMD-VMD adjacent reconstruction and secondary decomposition

Traditional empirical mode decomposition (EMD) algorithm has drawback of mode mixing when dealing with complicative vibration signals, and this will reduce the accuracy of bearing fault diagnosis. By considering the superior performance of variational mode decomposition (VMD) in signal processing, this paper proposes a bearing fault diagnosis method based on EMD-VMD adjacent reconstruction and secondary decomposition algorithm. Firstly, the vibration signal is decomposed by EMD. Secondly, a mixed index normalized from kurtosis and root mean square is proposed to locate the sensitive IMF component. Then, an adjacent reconstruction method is designed by combining the obtained sensitive component and its neighbor IMFs. Finally, the modal component parameter of VMD is set by these IMFs used for new reconstructed signal, and bearing fault diagnosis is realized with Hilbert analysis. Simulation analysis experimental analysis results show that the proposed method has advantages in baring fault diagnosis compared with some traditional methods.