In order to solve the problem of rolling bearing fault feature extraction in strong background noise environment,a rolling bearing fault diagnosis method based on HOA-TVFEMD is proposed.The optimal parameter combination is determined by the walking optimization algorithm,and the intrinsic mode function(IMF)is obtained by TVFEMD decomposition of the parameters.The component with the highest fault feature energy ratio is selected in the envelope spectrum,and the fault feature is extracted by envelope demodulation analysis for fault diagnosis.This method can identify the subtle fault characteristics of bearings and has high engineering application value.