A new de-noising method based on parameter optimized Mexican hat wavelet is put forward in this paper. For the similar shape to the mechanical shock vibration signals, the Mexican hat wavelet is chosen as the mother wavelet and improved by the shape parameters optimization. The noise jamming in the raw vibration signals can be filtered by the continue wavelet transform (CWT) using the improved Mexican hat wavelet as the mother wavelet. The shape parameters of the Mexican hat wavelet are optimized by the cross validation method (CVM). In the CWT process, the optimal scale factor is also obtained by the circle CVM. The useful components can be extracted by  the CWT with the optimal shape parameters and scale factor. The experimental result shows that the proposed method can not only de-noise the useless noise effectively, but also extracts the fault feature availably.