A Multi-channel Signal Denoising Method Based on Signal Correlation

Abstract

The invention discloses a multi-channel signal denoising method based on signal correlation and relates to a sparse decomposition denoising method for a multi-channel signal. The problems that the sparse decomposition method is low in denoising signal to noise ratio, the multi-path signal combined denoising cannot be realized and the signal characteristics and the signal correlation are not considered so that each original signal cannot be efficiently and accurately extracted from the acquired multi-channel signals in the conventional signal denoising field are solved. The multi-channel signal denoising method based on signal correlation mainly comprises the following steps: setting an initial parameter value; acquiring the most matching atoms of the residual error in the multi-channel signal set; acquiring a sparse decomposition matching sub-dictionary of the multi-channel signal set; acquiring the atomic number of the sparse decomposition matching sub-dictionary; updating the residual error of the multi-channel signal set; judging whether the iterations is small than the preset maximum iterations; estimating a sparse decomposition coefficient vector of the multi-channel signal set; and synthesizing the denoised multi-channel signal set. The method can be widely applied to noise elimination and suppression of multi-channel combined sparse signals with the signal correlation.

Description

technical field [0001] The invention relates to a method for sparsely decomposing and denoising multi-channel signals, which belongs to the field of signal noise elimination and suppression. Background technique [0002] The purpose of signal denoising is to discard all kinds of interference from the noisy data and extract the desired signal, which provides a strong guarantee for revealing the unknown information hidden in the signal. Decades of development have made some achievements in the research on signal noise elimination and suppression theory and its algorithm, and a variety of denoising methods have emerged, mainly including traditional filtering method, Wiener, Kalman filtering method, and SVD decomposition method. , wavelet decomposition method, empirical mode decomposition method, independent component analysis method, neural network and sparse decomposition and other methods. However, most of the different denoising methods are effective for specific signals an...

AI Technical Summary

Problems solved by technology

[0004] In order to solve the problem that the sparse decomposition method in the existing signal denoising field has a low denoising signal-to-noise ratio, joint denoising of multi-channel signals cannot be realized, and the characteristics of the signal itself and signal correlation are not considered, so that it is impossible to obtain high-efficiency data from the acquired multi-channel signals. Accurately extract the problem of each original signal, so as to propose a multi-channel signal denoising method based on signal correlation

Images