Abstract: In order to detect the rail corrugation fault effectively, this paper proposed a method of rail corrugation identification based on parameter optimization Variational Mode Decomposition (VMD) and Smooth Pseudo Wigner Ville Distribution (SPWVD). The paper used the VSSLMS (Variable Step Size Least Mean Square) algorithm to filter the original acceleration signal of train axle box vibration, decomposed the filtered signal into variational mode, took the envelope entropy of the decomposed signal as the index of the identification of the rail corrugation, used the SPWVD to carry out the time-frequency analysis of decomposed signal, and determine the location and wavelength of the corrugation. The effectiveness of the method was verified by simulation signals and examples. The results show that the method can be used to improve the accuracy of rail corrugation identification and assist rail maintenance.