一种自相关函数绝对值均值变点的去噪方法

A method of CEEMD denoising based on the theory of absolute value of autocorrelation function mean point variation

针对CEEMD去噪基于能量准则判定分界点存在稳定性差的问题,该文以噪声和有用信号各自的自相关函数特点之间的差异为基础,提出了一种基于自相关函数绝对值均值变点的CEEMD去噪方法。该方法利用噪声和有用信号各自的自相关函数值存在差异的特点,实现了噪声和信号模态分界点的定量判定,对发现的噪声模态分量,采用小波软阈值法去噪,保留了存在于噪声模态分量中有用的信号。bjfs和auck两个CORS站20余年的高程时间序列数据,实验结果表明:该方法可以定量判定噪声和信号模态的分界点,去噪后信噪比提高14%以上,而且采用均方根误差指标统计的高程时间序列预报精度提高了44%以上。提出了一种基于自相关函数绝对值均值变点的CEEMD去噪方法,实现了CORS站高程坐标分量数据的去噪。

According to the fact that the problem of poor stability in the CEEMD denoising based on energy criterion,a CEEMD denoising method based on the absolute value of the autocorrelation function is proposed,basing on the difference between the features of the autocorrelation function of noise and the useful signal.This method makes use of the difference of the value of autocorrelation function between noise and useful signal and realizes the quantitative determination of the boundary point of noise and signal mode.The wavelet soft threshold method is used to remove noise from the noise modal components.The experimental results of the elevation time series data of two CORS stations of BJFS and auck over 20years show that:This method can quantitatively determine the demarcation point of noise and signal mode,the signal-to-noise ratio increased by more than 14%after de-noising.,Moreover,the accuracy of the altimetric time series forecast using root mean square error index is improved by more than 44%.A method of CEEMD denoising based on the absolute value of autocorrelation function is proposed,realizing the denoising of the elevation coordinate component data of CORS station.