Independent Component Extraction Of Gravity Tidal Signals And Analysis Of Seismic Precursor Information

The Gravity Earth tide is a regular and periodic tidal phenomenon produced by the interaction of the earth and the sun and the earth's rotation.It is the only physical phenomenon that can be calculated.There are abundant geophysical information in the tide signal,which is widely used in geodesy,astronomy,geophysics and seismology.This paper aims at the earthquake precursory information contains gravity tidal in the signal.The gravity earth tide signal is composed of a plurality of independent tidal harmonic waves,and each independent harmonic is extracted according to the generating mechanism of the gravity tide signal.The independent component analysis(ICA)method of blind signal processing is used to decompose and extract the harmonic component of gravity tide signal.The classical ICA algorithm is combined with the intelligent optimization algorithm to overcome the shortcomings of the original ICA algorithm,and an improved independent component analysis method is proposed.In this paper,we use the genetic algorithm as the optimization algorithm,which is fast and simple.It is hoped that the analysis of the gravity tide signal can be obtained by using the harmonic components.Based on the previous results,this paper will use a three-dimensional orthogonal model,this model is based on the tidal gravity signal generating mechanism,the main impact will be the tidal gravity signal generated by the rotation of the earth,the moon and the sun of the tidal force were decomposed into three mutually orthogonal directions,using this model which factors are the harmonics analysis of the extracted.Finally,the genetic algorithm of independent component analysis method and 3D orthogonal analysis through the closed formula model of tidal gravity signals on the use of the algorithm to obtain the calculated to the theoretical value of signal separation based on the independent.The results show that the proposed method and the analytical model are feasible and effective.Then add the measured signal to the experiment,and analyze the seismic precursor information of each component.