In order to solve the problem that the noise of the hammer powersignal in mold repair is large

the dbN wavelet (abbreviation of Daubechies

Nis the wavelet order) is used to reconstruct the hammer force signal accordingto the obtained low frequency coefficient. The low-frequency coefficients of thesignal are reserved and the high-frequency coefficients are discarded Ultimately

the signal-to-noiseratio after noise reduction is 10.0778and the mean square error is 0.6633

the noise reduction of hammer power signal in mold repair isinitially achieved. Simultaneously

the thresholding is done on the basis ofdbN wavelet decomposition. Experimental results show that the SNR issignificantly increased after threshold treatment

up to 39.85dB; the RMSE issignificantly reduced and the minimum value is 0.4498. Comprehensive macro-waveformcharacteristics indicating that the two methods can achieve noise filtering ofhammering force signal. Besides

soft threshold methods can also be a greaterdegree of reduction of the original signal at the mutation point for thedetails characteristics

to avoid the signal distortion and ensure that thesubsequent calculation accuracy of hammering force signal in mold repair.