Research On Adaptive Denoising Method For Chaotic Signals

Chaotic theory has been widely applied in the fields of secure communication,biomedicine,chemical production and so on.However,in practical applications,the measured chaotic signals are usually contaminated by noise due to various subjective and objective factors.The existence of noise makes the characteristics of chaotic signals obscured,which makes the study of chaotic theory and the application of chaotic theory are limited in various fields.Therefore,research on chaotic signal noise suppression has important applicable value and theoretical significance.In this paper,the chaotic signal denoising problem is deeply studied,and proposes some solution to solve the problem in the existing denoising methods.The specific content is summarized as follows:(1)In order to solve the parameter optimization problem in the denoising method for chaotic signal based on collaborative filtering,considering the selection of optimal filtering parameters is affected by signal characteristics,sampling frequency and noise level,a chaotic signal adaptive filtering noise suppression based on permutation entropy is proposed.Firstly,different parameters are used for denoising noisy chaotic signal,then the entropy algorithm is used to characterize the complex of denoised signal,and the parameter corresponding to the minimum permutation entropy is selected as the optimal parameter.Simulation results show that the proposed method can automatically optimize each filter parameter under different signal characteristics,sampling frequency and noise level,which improves the adaptivity of the denoising method for chaotic signals based on collaborative filtering.(2)In order to solve the mode mixing phenomenon,difficulty in threshold selection and limited signal-to-noise ratio enhancement,a zero-crossing scale threshold denoising method based on improved EMD is proposed.This method uses the zero-crossing scale threshold and the Durbin-Watson(DW)statistic to analyze the correlation of denoised signal residuals to find the optimal threshold,which improves the adaptivity of this denoising method.Firstly,the improved EMD is used to decompose the noisy signal.Then the different thresholds are used for denoising chaotic signal,and calculate the DW value of residual signal.The threshold corresponding to the DW value of signal residual which is the closest to the DW value of gaussian white noise is selected as the optimal threshold.The final denoising signal is obtained by using the optimal threshold.The chaotic signal,ECG signal and sunspot signal are used as test signals to verify the performance of the proposed method.Simulation results show that the proposed method effectively solves the mixing phenomenon and difficulty in threshold selection in the interval threshold denoising method based on empirical mode decomposition,which achieve higher adaptivity and signal to noise ratio.