Fault Detection of Bearing in Induction Motor Using Improved Variational Nonlinear Chirp Mode Decomposition

Abstract

When the motor bearing fails, the weak fault feature in the stator current signal are submerged in the strong noise background of fundamental and harmonics. Due to the low signal-to-noise ratio, bearing fault detection based on current is always a challenge. Although the variational nonlinear chirp mode decomposition (VNCMD) can process non-stationary signals, it requires some prior conditions, which limits its practical application. For this, an improved VNCMD is proposed for the current-based bearing fault detection. Firstly, the initial instantaneous frequency is estimated based on Fourier series. Secondly, the current signal is decomposed into multiple modes based on the VNCMD using the initial instantaneous frequency as initial conditions. Finally, the most relevant mode is selected using the correlation coefficient method. The experimental results show that the proposed approach is effective for the bearing fault detection, and has superior performance comparing with the existing approaches.