基于神经网络的地震震相自动拾取方法  

Neural-network-based seismic phase automatic pickup method

作　　者：姚开一 李英玉 YAO Kai-yi;LI Ying-yu(National Space Science Center,Chinese Academy of Sciences,Beijing 100190,China;University of Chinese Academy of Sciences,Beijing 100049,China)

机构地区：[1]中国科学院国家空间科学中心,北京100190 [2]中国科学院大学,北京100049

出　　处：《电子设计工程》2018年第22期1-5,共5页Electronic Design Engineering

基　　金：国家863计划资助项目(2015AA7013040);中科院重点部署项目(ZKRW-KT2016-2)

摘　　要：地震预警的基本原理是利用P波比S波速度快的特性,对先至的P波进行快速拾取和分析,以此实现对破坏性较大的S波的提前预警,减少人员伤亡和财产损失,因此地震波震相的高速拾取便显得至关重要。传统的地震震相自动拾取方法费时费力,且误差较大,因此研究和发展新的拾取方法和系统意义十分重大。影响地震识别的因素十分复杂,本课题尝试将LSTM神经网络应用于地震信号处理,利用神经网络的高度非曲线映射功能更好的总结地震识别规律,实现更加高效、精确的P波自动拾取方法。The basic principle of earthquake early warning is to use the characteristics that P wave spreads faster than S wave,quickly pick up and analyze the first arrived P wave so as to prevent the Serious destruction of S wave and reduce the casualties and property losses,So the high-speed pick-up of the seismic P wave phase is very important.The traditional method of picking up the phase of earthquakes automatically takes time and effort,and the error is so large.The factors affecting the seismic identification are very complicated.The subject attempts to apply LSTM neural network to seismic signal processing,and use the highly non-curve mapping function of neural network to summarize the law of seismic identification to achieve a more efficient and accurate P wave automatic pick-up method.

关 键 词：震相拾取 P波S波拾取 神经网络 LSTM 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]