In view of the RV reducer has many vibration sources, complex vibration signals, and it is difficult to extract the internal weak fault characteristic information effectively, a weak fault diagnosis method based on the combination of variational mode decomposition and S-transform marginal spectrum (VMD-STMS) was proposed in this paper. Firstly, the signals were decomposed and de-noised by means of variational modal decomposition, and the eigenmodal functions with different center frequencies were obtained. Then the kurtosis index of the intrinsic mode function is calculated, and the IMF component with the largest kurtosis value is selected as the optimal component. Finally, the S-transform time-domain marginal spectrum method is used to identify the weak faults from the optimal components. The validity of the method is verified by analyzing and processing the weak fault data of RV reducer, such as crack and pitting of planetary gears. The analysis results accord with the theoretical value of fault frequency, and the fault type is diagnosed successfully.
 

RV reducer; Variational mode decomposition; S transform; Marginal time spectrum; Weak fault diagnosis