For the randomness and blindness of removal of the noisy component and false component after local wave decomposition

a method of local wave mutual information noise reduction is proposed.Firstly

the decomposition of the signal is carried out by using the local wave and the mutual information of each IMF component with the original signal is calculated.Secondly

the component with high mutual information is selected for autocorrelation analysis to remove the noise component

and the signal is reconstructed by using the valid component.In order to verify the effectiveness of the method

the simulation signal is used for analyzing

and applied it to the gearbox fault diagnosis.The gearbox vibration signal wavelet transform is carried out after local wave mutual information noise reduction

and the wavelet singular spectrum entropy is extracted as the fault feature to identify faults.The results are compared with the noise reduction processing recognition results prove the practicality of the local wave mutual information noise reduction in engineering practice.