利用天然地震震源和人工爆破震源之间信号能量分布的差异，结合RBF神经网络技术，对2类事件进行分类，具体步骤如下：使用8个带通滤波器对事件波形进行滤波，并划分为4个波形段：P波、P波尾波、S波和S波尾波，分别计算每个滤波器信道和波形段的能量特征值，以所得32个特征参数作为输入向量，利用RBF神经网络，对地震和爆破事件进行分类识别。结果表明，基于RBF神经网络的地震事件识别方法，识别率为88.1%，具有较高的准确性，可作为地震与爆破事件识别的一个重要依据。
In this paper, eight band-pass filters are used to filter the events waveform, which are divided into four waveform segments including P, P coda, S, and S coda. Then, the energy eigenvalues of each filter channel and waveform segment are calculated respectively. Finally, taking the 32 characteristic parameters as input vectors, the RBF neural network is used to classify and identify earthquake and explosion events. The results show that the recognition rate of the method based on the RBF neural network is 88.1%, which has high accuracy and can be used as an important basis for earthquake and blasting event identification.
2020,41(3): 59-66 收稿日期：2020-01-09
DOI：10.3969/j.issn.1003-3246.2020.03.008
基金项目：广西科学研究与技术开发计划（项目编号：桂科AB1850042）；红水河流域水库地震特征的精细研究——以天峨至大化段为例；广西壮族自治区地震局科研课题（项目编号：2017002）；人工神经网络技术在爆破识别中的应用
作者简介：毛世榕(1984-),男,工程师,主要从事地震监测工作。E-mail:232529840@qq.com
参考文献：
毕明霞,黄汉明,边银菊,李锐,陈银燕,赵静. 天然地震与人工爆破波形信号HHT特征提取和SVM识别研究[J]. 地球物理学进展,2011,26(4):1 157-1 164.
边银菊,王婷婷,郭永霞. 用决策方法识别地震与爆破[J]. 地震学报,2012,34(3):397-407.
范晓易,曲均浩,曲保安,刘方斌,山长仑,周少辉. 支持向量分类机LIBSVM方法识别天然地震、爆破与塌陷[J]. 大地测量与地球动力学,2019,39(9):916-918.
黄汉明,边银菊,卢世军,蒋正锋,李锐. 天然地震与人工爆破的波形小波特征研究[J]. 地震学报,2010,32(3):270-276.
靳玉贞,林木金,范晓瑜,刘晓萍,何佳,杨世英,孟彩菊. 山西地区爆破、塌陷(矿震)特殊地震动特征识别[J]. 地震地磁观测与研究,2015,36(3):63-66.
王婷婷,边银菊,张博. 地震与爆破的小波包识别判据研究[J]. 地震学报,2014,36(2):220-232.
王婷婷,边银菊. 振幅衰减特性在地震与爆破识别中的应用[J]. 地震学报,2015,37(1):169-179.
杨芳,朱嘉健,刘智,吴彬. 广东地区地震与爆破事件识别方法研究[J]. 华南地震,2016,36(3):110-115.
张博,边银菊,王婷婷. 用逐步代价最小决策法识别地震与爆破[J]. 地震学报,2014,36(2):233-243.
张萍,蒋秀琴,苗春兰,邹向荣. 爆破、矿震与地震的波谱差异[J]. 地震地磁观测与研究,2005,26(3):24-34.
中国地震局监测预报司. 测震学原理与方法[M]. 北京:地震出版社,2017:283-284.
Beccar-Varela M P, Gonzalez-Huizar H, Mariani M C, Tweneboah O K. Use of wavelets techniques to discriminate between explosions and natural earthquakes[J]. Physica A:Statistical Mechanics and its Applications, 2016, 457:42-51.
Fäh D, Koch K. Discrimination between earthquakes and chemical explosions by multivariate statistical analysis:a case study for Switzerland[J]. Bulletin of the Seismological Society of America, 2002, 92(5):1 795-1 805.
Haykin S. Neural networks:a comprehensive foundation[M]. Beijing:Tsinghua University Press, 2001:278.
Kuyuk H S, Yildirim E, Dogan E, Horasan G. Application of k-means and Gaussian mixture model for classification of seismic activities in Istanbul[J]. Nonlinear Process Geophys, 2012, 19(4):411-419.
Kuyuk H S, Yildirim E, Dogan E, Horasan G. Clustering seismic activities using linear and nonlinear discriminant analysis[J]. Journal of Earth Science, 2014, 25(1):140-145.
Lyubushin A A, Kaláb A, Lednická Z M, Haggag H M. Discrimination of earthquakes and explosions using multi-fractal singularity spectrums properties[J]. Journal of Seismology, 2013, 17(1):975-983.
Pasyanos M E. A general method to estimate earthquake moment and magnitude using regional phase amplitudes[J]. Bulletin of the Seismological Society of America, 2010, 100(4):1 724-1 732.
Yılmaz Ş, Bayrak Y, Cinar H. Discrimination of earthquakes and quarry blasts in the eastern Black Sea region of Turkey[J]. Journal of Seismology, 2013, 17(2):721-734.
