Study On Seismic Event Location Based On The Combinations Of Waveform Features And Bayesian Detection

The acquisition of seismic signal can be easily influenced by many kinds of factors due to the complex structure of the earth itself,which will also affect the accuracy of seismic event detection.The existing seismic event location model based on multi-steps processing of seismic signals may lose some potential information carried by the waveforms.Therefore,improving the information utilization and the reliability of the model becomes a hot topic recently.Based on the study of the existing location method,this paper enhances the method based on Bayesian detection model as follows:In this paper,a single-station seismic event location method based on the combination of waveform features and Bayesian detection model is proposed to solve the problem of information loss in existing model.Compared with the existing location method,not only the features of seismic data obtained from existing multi-steps processing methods,but also the envelope features of the seismic signal are extracted for modeling.In order to obtain the features of seismic signal waveform,the envelopes of the waveform are parameterized at first,then some important features are extracted to build the model,such as the rise time,maximum amplitude and decay rate of the waveform.In this paper,the grid search method is used to grid the geographic area to be detected by the longitude and latitude of the earth,and all grid points are searched and tested for locating the earthquake.The location method proposed in this paper makes full use of the information embedded in the waveform and improves the reliability and instantaneity of location method.To tackle the problem of the large error and instability which exist in the single-station seismic event location method,a joint multi-station seismic event location method based on the combination of waveform features and Bayesian detection model is proposed.In this paper,the single-station method is used to find the area of the event roughly,and the joint multi-station method is used to find the accurate longitude and latitude of the event,so as to complete the location gradually.Compared with the single-station location method,the joint multi-station location method takes fully consideration of the difference of detection information between multiple stations,further reduces the error and helps to improve the accuracy of seismic location.With the proposed location method in this paper,a seismic event location test platform is built by using Python,and the effectiveness of the method is verified on this platform.The test results show that the location method proposed in this paper have small error and can effectively locate seismic events.