论文部分内容阅读
分形信号中Hurst指数的提取是一个十分重要的研究方向 .为此提出了一种新的基于小波变换和分数高斯噪声模型的Hurst指数提取方法 .分别推导了分形信号和白噪声的小波系数统计量的公式 ,在此基础上用最大似然法进行参数估计 .与传统方法的对比实验表明 ,该方法具有较高的稳定性
The extraction of Hurst exponent in fractal signal is a very important research direction.Therefore, a new method of Hurst exponent extraction based on wavelet transform and fractional Gaussian noise model is proposed.Wavelet coefficient statistics of fractal signal and white noise Based on which the maximum likelihood method is used to estimate the parameters.Comparison with the traditional method shows that this method has high stability