井下指令信号的小波-奇异值分解双层滤波降噪

Double-layer Filtering Denoising of Downhole Command Signal Based on Wavelet Transform-Singular Value Decomposition

针对旋转导向钻井井下接收的指令信号进行降噪时存在奇异点和突变点等问题,提出一种基于小波-奇异值分解的双层滤波降噪方法。该方法将小波改进分层阈值降噪与奇异值分解降噪相结合,采用小波变换与奇异值分解WT-SVD双层滤波降噪方法,对小波改进分层阈值降噪后的信号进行矩阵构造,经过奇异值分解和重构使信号得以还原。通过实验确定了所选取的奇异值个数及重构矩阵结构,并对比研究了奇异值分解、小波改进分层阈值和WT-SVD双层滤波3种不同降噪方法降噪后的信号波形及信噪比、均方根误差和相关系数。实验结果表明,WT-SVD双层滤波方法降噪效果明显。

Singular and abrupt points exist in the noise reduction of the command signals received downhole in rotary steering drilling.In order to solve this problem,a two-layer filtering denoising method based on wavelet transform-singular value decomposition is proposed.This method constructs the matrix of the signal after the wavelet-improved hierarchical threshold denoising by combining the wavelet-improved hierarchical threshold denoising with singular value decomposition denoising and using WT-SVD(Wavelet Transform-Singular Value Decomposition)double-layer filtering denoising method and restores the signal through singular value decomposition and reconstruction.The number of selected singular values and the structure of reconstructed matrix are determined through experiments.The signal waveform,signal-to-noise ratio,root mean square error and correlation coefficient of signal after denoising using three different noise reduction methods,namely,singular value decomposition,wavelet-improved hierarchical threshold and WT-SVD double-layer filtering,are compared and studied.The result shows that WT-SVD double-layer filtering method has better noise reduction effect than other two methods.