Research On Recognition Method Of Life Signal In Ruins After Earthquake Based On Blind Source Separation

In recent years,our country's seismic activities have been frequent,and strong vibrations have caused disasters such as building collapse,ground damage,and landslides.When the earthquake broke out,some people were buried in the rubble because they could not escape,and even some people who were buried could not be found in time and missed the best rescue opportunity,causing second casualties.In the complex environment of the ruins after the earthquake,how to quickly and efficiently identify the characteristic signals of life bodies is particularly important for accurate rescue of the trapped persons.Blind source separation technology has been widely used in biomedicine,signal processing,mechanical fault diagnosis and other fields,including the research of EEG signal separation,voice signal separation,gear fault diagnosis,etc.,when the source signal and signal mixing method are unknown,The source signal required for separation can be identified by observing the statistical characteristics of the signal.In this paper,blind source separation technology is applied to the identification of life signals in ruins after the earthquake.Independent Component Analysis(ICA)algorithm and Fast Independent Component Analysis(Fast Independent Component Analysis,Fast ICA)algorithm are used to deal with multi-sources in complex ruin environment.Mixed signals,identification and separation of vital signs.The pros and cons of the algorithm can be evaluated in terms of its convergence speed and stability.Different algorithms directly determine the separation effect.The ICA algorithm and Fast ICA algorithm in the blind source separation method are used to study the signal separation effect of noisy sinusoidal signals,square wave signals,sawtooth signals,and random signals,and the effectiveness of the algorithm is verified by experiments in Matlab.Using two performance evaluation indicators,signal-to-noise ratio and running time,the pros and cons of the signal separation effects of ICA and Fast ICA are compared.The experimental results show that the average signal-to-noise ratio of the ICA algorithm is 2.4052 d B,the average time used for one run is 5.4049 s,the average signal-to-noise ratio of the Fast ICA algorithm is 6.2838 d B,and the average time used for one run is 0.0223 s.The Fast ICA algorithm Better signal separation effect.In order to further verify the feasibility of the Fast ICA algorithm in the identification of life signals in the ruins after the earthquake,the life signature signals in the environment of noisy human voices and large-scale mechanical equipment operations are collected,and the Fast ICA algorithm is used to conduct identification experiments and analysis of different environmental signals,and use evaluation The index analysis separation effect,the experimental results show that the minimum correlation coefficient of Fast ICA algorithm is 0.9974,and the longest running time is 0.1125 s,which verifies the feasibility of using Fast ICA algorithm to identify and separate vital signs in the complex environment of post-earthquake ruin rescue.