Design Of Intelligent Alarm Terminal Based On Bayesian Seismic Event Discrimination

As a means of disaster prevention and mitigation,seismic alarm can judge the occurrence and harm degree of earthquake by monitoring seismic wave,and send out alarm notice in time.Its role has been widely recognized and received more and more attention.If a single station-based seismic alarm can be installed in a densely populated area or in an important facility,it has a positive effect on reducing casualties and property losses.Among the existing ground motion prediction methods of a single station,the following ground motion parameters are estimated based on the initial P wave information recorded by the station,which inevitably leads to a large prediction error.It is difficult to distinguish seismic events,such as small near-earthquake or large-earthquake far-off earthquake.Based on Bayesian theory,the magnitude and PGA prediction method is studied in this paper,which can obtain more accurate magnitude and PGA valuation,and complete the division of seismic events with four level judgment method,which can effectively improve the accuracy and pertinence of single station earthquake alarm.The seismic data of Wenchuan and Jiuzhaigou are first selected,and the seismic phase of P waves is picked up automatically by combining STA\LTA and AIC criteria.then,the relationship between the characteristic period and the magnitude fitting within 3 seconds after the arrival of the P wave,and the fitting relationship between the displacement amplitude and the PGA counted respectively.A real-time prediction method of magnitude and PGA based on Bayesian theory is proposed.The error analysis of the prediction results of the two prediction methods is carried out.The comparison shows that the prediction method based on Bayesian theory can effectively reduce the prediction error and provide more accurate warning information for the earthquake warning terminal.A STM32F103C8T6 controller is designed to control four different colorful and acoustooptical alarms in real time by using the minimum system.The alarm has the characteristics of low cost,easy operation and strong applicability.Finally,an example of the June 17,2019 Sichuan Yibin 6.0 earthquake is given to verify.The experimental results show that the alarm terminal based on Bayesian seismic event discrimination can realize the classification of seismic events and can send out alarm messages more accurately and effectively.