针对混沌信号和噪声频谱互相重叠,传统方法难以实现有效滤波这一问题,提出一种改进的小波去噪方法.该方法采用参数加权法构造信号,将小波分解系数进行阈值处理,通过循环迭代,利用序列中包含的有效信息,将有用信号提取出来.仿真结果表明,利用改进小波变换去噪方法改善了混沌时间序列的预测结果,证明了该方法的有效性.
The traditional filtering method can not reduce the noise in the chaotic system effectively as the frequency spectrum of the chaotic signals and noise are overlapped.In order to solve this problem,an improved wavelet denoising method based on weighted parameters method is proposed.The noisereduction idea is that reconstruct signals after the threshold compromise of the wavelet decomposition.Then the useful signals are extracted from the original chaotic signals and the reconstructed signals.Simulation results show that the improved wavelet denoising method improves the effect of chaotic forecast.