In order to solve the problem that the early fault vibration signal of rolling bearing has low signal-tonoise ratio and difficult feature extraction

a fault feature extraction algorithm based on variational mode decomposition（ VMD） and morphological difference filter is proposed. VMD of the fault signal is carried out to obtain the intrinsic mode function（ IMF）. kernel component with the largest kurtosis value is selected by the kurtosis criterion

and the feature extraction is carried out. The signal is envelope spectrum analysis

and fault characteristic frequency of signal is obtained. The method is applied to the data collected by the rolling bearing.The results show that the method improves decomposition efficiency of signal and reduce influence of noise

and can realize accurate diagnosis of the rolling bearing fault.