A Combined Signal Denoising Method Based on Empirical Mode Decomposition and Wavelet Analysis

Abstract

The invention belongs to the technical field of signal processing, in particular to a denoising method for additive Gaussian white noise signals under low signal-to-noise ratio. According to the autocorrelation of the signal, the present invention calculates the autocorrelation of the signal, the autocorrelation function of the signal obtains the maximum value at the zero point, and the amplitude changes with the change of the time difference, and will not quickly decay to a small value . EMD decomposition is performed on the signal mixed with Gaussian white noise. Due to the nature of EMD decomposition, Gaussian white noise is no longer true white noise, but the statistical properties of white noise approximately exist, that is, the signal of Gaussian white noise mixed with The autocorrelation function achieves the maximum value at zero point, and the amplitude changes with the time difference, but it decays quickly with time. This difference can be used to select the IMF component in which the noise plays a dominant role and effectively reduce the influence of noise on the signal. Under the condition of low signal-to-noise ratio, the denoising performance of the invention is better than that of the traditional method, and the signal denoising can be completed under the condition of low signal-to-noise ratio.

Description

technical field [0001] The invention belongs to the technical field of signal processing, in particular to a denoising method for additive Gaussian white noise signals under low signal-to-noise ratio. Background technique [0002] Signals usually contain a lot of useful information, such as frequency characteristics, time characteristics, etc. The purpose of signal analysis is to transform some information characteristics of the signal into a form that is easy for people to understand by certain means, so as to better understand the physical characteristics represented by the signal. In order to extract useful information from the signal, people have studied a variety of transformation and analysis methods in order to better observe and analyze the signal from multiple angles. Most of the traditional signal processing methods are based on the assumption of linear and stable Gaussian signals, but many signals in real life and production work are mostly nonlinear and non-stat...

AI Technical Summary

Problems solved by technology

When the signal-to-noise ratio is low, since the useful information energy of the signal is very small, and the noise part becomes smaller with the increase of the number of decomposition layers, the threshold should be adaptive. Poor denoising performance compared to

Images