Chaotic Signal Denoising Based on Threshold Selection of Wavelet Transform

A new method is proposed to reduce noise within chaotic signal based on multi-resolution wavelet transform.This method can effectively extract the nonlinear signals by decomposing the noisy details with wavelet(analysis),and can choose different thresholds neatly according to the signal noise ratio on different scales.This method is more accordant to the application and can improve the noise reduction results.Chaotic time series generated by Lorenz system and the Yellow River annual runoff time series are simulated to compare with other methods,and the results of simulation prove the effectiveness of the proposed method.