CEEMD denoising signal reconstruction method and device based on permutation entropy accurate measurement

Abstract

The invention discloses a CEEMD denoising signal reconstruction method and device based on permutation entropy accurate measurement, and the method comprises the steps: S1, preprocessing an original signal, so as to improve the signal-to-noise ratio of echo data; S2, performing CEEMD reconstruction; using improved CEEMD decomposition to obtain IMF components, determining an IMF demarcation point through permutation entropy and a maximum between-class variance method, and marking the IMF demarcation point and the IMF components behind the IMF demarcation point as useful components; S3, reconstructing a signal; and performing signal reconstruction by using the useful component after CEEMD reconstruction. The device is used for implementing the method. The invention has the advantages of being simple in principle, better in effect, wider in application range, capable of improving the signal denoising reconstruction effect and the like.

Description

technical field [0001] The invention mainly relates to the technical field of ultra-wideband radar, in particular to a CEEMD denoising and reconstructing signal method and device based on permutation entropy quasi-measurement. Background technique [0002] UWB radar is a kind of detection radar with strong penetrability, high resolution and all-weather performance. After years of development, it is widely used in the fields of anti-terrorism in street fighting, disaster rescue and medical monitoring. In practical applications, although ultra-wideband radar can detect human breathing, heartbeat and other signals, most of them are distributed in low-frequency segments, but traditional low-pass filters and wavelet transforms can retain most of the target signals, thereby suppressing noise; However, the existing methods still have problems such as smoothing mutation signals and lack of adaptability, which further affect the effect of dealing with noise. [0003] Since Empirical...

AI Technical Summary

Problems solved by technology

[0004] However, the above EMD method also has some problems: in decomposing signals, it is easy for signals of similar scale to appear in different IMF components or a single IMF contains signals of different scales, that is, the problem of mode aliasing, which not only affects the noise Suppressive effects have a certain impact and also produce false target signals

[0006] 1. Adding white noise requires manual settings;

[0007] 2. The number of times of adding white noise is on the order of hundreds, resulting in a long time-consuming;

[0008] 3. The IMF obtained by the average processing does not meet the requirements, and there is still white noise after signal reconstruction, etc.

[0013] (1) The EEMD method itself has the problems of long time consumption and residual noise after signal reconstruction;

[0014] (2) The method of using preset thresholds to screen IMF components is not universal

Images