n tracked vehicle vibration signal data processing, the existence of outliers can produce false harmonic frequency components. By using wavelet transformation vibration signal is decomposed into approximate components and details components. With the Wright standards and the characteristics of outlier’s wavelet coefficients, outlier’s details components are replaced by new details components, which are calculated by weighting coefficient method. The outliers are removed and signal is reconstructed. Test data processing results show that the new method can maximum retain the useful signal information, realizing the outliers eliminate.

n order to avoid an interference of strong noise to tracked vehicles vibration signal, using wavelet transform, several wavelet bases was compared according to reconstruction factors, then an appropriate wavelet base and threshold to processing vibration signals of tracked vehicles were determined, the wavelet layered method was described The simulation result indicated the wavelet transformation de-noising method can suppress the background noise effectively, extract useful state information, enhance the signal-to-noise ratio, reduce mean-square error and recover the interest weak signal buried in intensive background noise.