Font Size: a A A

EDA-based ICA Method And Its Application In Geophysical Information Analysis Of Gravity Solid Tides

Posted on:2018-03-12Degree:MasterType:Thesis
Country:ChinaCandidate:A Y ZhangFull Text:PDF
GTID:2350330518960464Subject:Electronics and Communications Engineering
Abstract/Summary:PDF Full Text Request
With the development of the instrument industry,and the many studies of the earth,more and more scientists think that the earthquake can be predicted.And then,how to get the seismic precursor information becomes more and more popular in academic circle.Analysis the earth tide is a kind of methods to get the precursor information.The gravity earth tide is one kind of the earth tide to use to get the changing of the gravity,which can be used to get the interior structure and its future motion.The gravity earth tide contains many harmonic component,such as the long-period waves,diurnal wave,and semidiurnal wave,which are all caused by the Moon,the Sun and the earth rotation.The key to get the geophysical information or the seismic precursor information is to analysis these harmonic component.An orthogonal decomposition model of gravity earth tide is proposed to decompose the gravity earth tide into three orthogonal direction,in order to extract the independent harmonic components from the gravity earth tide.In order to realize the extraction,his paper use ICA to verify and analysis.Independent Component Analysis(ICA)aims to separate the relatively independent signals from the mixed signal source.However,the original ICA is so sensitive to the initial values of the separation matrix,that affects the separation effect and even result in inconvergence if the initial values are not chosen appropriately.EDA(Estimation Distribution Algorithm)is a kind of optimization algorithm,which discards the steps of crossover and recombination in the genetic algorithm,and uses the probability model of the population to continuously update,so as to get the optimum solution.In this paper a new method based on EDA to combine ICA is formed,the learning model in EDA is used to optimize the separation matrix to solve the disadvantage.The simulation experiments show a better result to the original ICA.Then this paper uses the new method to separates gravity earth tide signal into three independent parts,which are long-period waves,diurnal wave,and semidiurnal wave.Every part represents the signal which is corresponding to the theory frequencies of the harmonic component in gravity earth tide signal.At the same time,the independent components are consistent with the orthogonal decomposition model of gravity earth tide signals,which correspondence to the orthogonal harmonic components in the model.In the end,the paper uses the new method to analysis of earthquake precursor information in gravity tide,which is used to analysis the three kinds of harmonic component.According to analysis the long-period wave to study the instantaneous sequence variation characteristics.The Yunnan earthquakes cases show that long-period wave exist abnormal variation,and the changes emerge 30 days before and after the earthquake.The change of the abnormal variation is easy to observe and analysis,which has the clear physical meaning to the analysis of the earthquake.
Keywords/Search Tags:gravity tide signal, independent components analysis, estimation distribution algorithm, seismic precursor information
PDF Full Text Request
Related items