基于多层全连接神经网络的6C地震波极化向量识别研究  

Six-Component Seismic Waves Polarization Vectors Identification Based on Multi-Layer Fully Connected Neural Network

作　　者：廖成旺[1,2,3] 庞聪 江勇[1,2,3] 吴涛 LIAO Chengwang;PANG Cong;JIANG Yong;WU Tao(Institute of Seismology,CEA,40 Hongshance Road,Wuhan 430071,China;Wuhan Gravitation and Solid Earth Tides,National Observation and Research Station,40 Hongshance Road,Wuhan 430071,China;Hubei Earthquake Agency,48 Hongshance Road,Wuhan 430071,China)

机构地区：[1]中国地震局地震研究所,武汉市430071 [2]武汉引力与固体潮国家野外科学观测研究站,武汉市430071 [3]湖北省地震局,武汉市430071

出　　处：《大地测量与地球动力学》2024年第4期331-335,435,共6页Journal of Geodesy and Geodynamics

基　　金：中国地震局地震研究所和应急管理部国家自然灾害防治研究院基本科研业务费(IS201916293,IS202236328);武汉引力与固体潮国家野外科学观测研究站开放基金(WHYWZ202208);中国地震局“三结合”课题(3JH-202201024)。

摘　　要：利用机器学习原理,提出一种基于多层全连接(multi-layer fully connected, MFC)神经网络的六分量(six-component, 6C)地震波极化向量识别方法。首先利用6C地震波各波型极化向量数学模型和一系列仿真参数得到5种波型和噪声波型各5 000个极化向量数据集,然后随机选取其中5 000个作为测试集,其余划分为训练集,进行MFC神经网络与支持向量机(support vector machine, SVM)的综合辨识性能对比实验。结果表明,MFC神经网络模型识别5种极化向量类型(SH波和Love波视为一类)和6种极化向量类型的效果均显著优于SVM模型,平均识别率分别达到99.786%和87.940%。Using principles of machine learning,we propose a six-component(6C)seismic waves polarization vector identification method based on multi-layer fully connected(MFC)neural network.Firstly,each 5000 polarization vector data sets for five wave types and noise are obtained by using the mathematical model of polarization vectors of 6C wave types under a series of simulation parameters.5000 of them are randomly selected as test sets and the others as training sets.We make a comprehensive comparison for identification performance between MFC neural network and support vector machine(SVM)model.The results show that the MFC neural network model is significantly better than the SVM model both in identifying five(SH and Love waves are treated one type)and six polarization vector types,with the average recognition rate of 99.786%and 87.940%,respectively.

关 键 词：极化向量识别 六分量地震波 多层全连接神经网络 支持向量机 

分 类 号：P315[天文地球—地震学]