Abstract: In order to improve the signal-to-noise ratio and the effectiveness of noise reduction, a redundant second generation wavelet transform based on interpolation principle was constructed. This method does not mesh in the process of decomposition, which the approximation signal and the detail signal have the same length with the original signal, so the amount of data is doubled the original signal. Data redundancy characteristics of the method has a good result of the analysis. Compared to the traditional wavelet transform and the second-generation wavelet transform, the redundancy of the second generation wavelet transform is more effective.