以神经网络预测地震为主题，采用长短时记忆(LSTM)神经网络构建地震预测模型。基于1932—2021年的南加州地震目录资料,数据按照0.8∶0.2的比例划分，训练集的时间窗口为1932年1月至2002年3月，测试集为2002年3月至2021年9月。模型以LSTM神经网络为核心，综合训练集地震时间序列数据中计算出的11个反映地震时空强度分布特征的地震活动性指标，以及与之对应的最大震级构建标签，对测试集进行回溯性预测检验。利用混淆矩阵中的准确率、精确度、R评分等指标评估模型预测效果。结果显示，该模型在地震预测方面有一定的成效，成功预测出2010年4月的7.2级大地震，并且部分模型的R评分高于我国目前的中期预测水平。然而仍有部分指标未达到理想状态，还需深入探讨。
This paper explores earthquake prediction using neural networks, focusing mainly on using long-short-time memory (LSTM) neural networks to construct an earthquake prediction model. Based on the Southern California earthquake catalog data from 1932 to 2021, the earthquake catalog from January 1932 to March 2002 was used as the training set(80% of the entire earthquake catalog), and the earthquake catalog from March 2002 to September 2021 was used as the test set (the remaining 20%). The LSTM neural network was selected, and 11 earthquake prediction factors reflecting the spatiotemporal intensity distribution characteristics of the earthquake time series data were calculated from the training set. The maximum magnitude label corresponding to these factors was used to construct the model. The test set was then used for retrospective prediction testing. The model’s prediction performance was evaluated using metrics such as accuracy, precision, and R-value, which were calculated based on the values in the confusion matrix. The results show that the prediction has achieved certain results, predicting the M7.2 earthquake in April 2010. The R-value of some models is significantly higher than China’s current medium-term prediction level. However, the value of the evaluation model is still not ideal, and further exploration is needed.
[1] Geller R J. Earthquake prediction: a critical review[J]. Geophysical Journal International,1997,131(3):425-450. DOI: 10.1111/j.1365-246X.1997.tb06588.x.
[2] Allen C R. Responsibilities in earthquake prediction[J]. Bulletin of the Seismological Society of America, 1976, 66(6): 2069-2074. DOI: 10.1785/bssa0660062069.
[3] Alves E I. Earthquake forecasting using neural networks: results and future work[J]. Nonlinear Dynamics, 2006, 44(1): 341-349. DOI: 10.1007/s11071-006-2018-1.
[4] Madahizadeh R, Allamehzadeh M. Prediction of aftershocks distribution using artificial neural networks and its application on the may 12, 2008, Sichuan earthquake[J]. Journal of Seismology and Earthquake Engineering,2009(3),11:112-120.
[5] Panakkat A, Adeli H. Neural network models for earthquake magnitude prediction using multiple seismicity indicators[J]. International Journal of Neural Systems,2007,17(1):13-33. DOI: 10.1142/S0129065707000890.
[6] Sri Lakshmi S, Tiwari R K. Model dissection from earthquake time series: a comparative analysis using modern non-linear forecasting and artificial neural network approaches[J]. Computers & Geosciences, 2009, 35(2): 191-204. DOI: 10.1016/j.cageo.2007.11.011.
[7] Negarestani A, Setayeshi S, Ghannadi-Maragheh M, et al. Layered neural networks based analysis of radon concentration and environmental parameters in earthquake prediction[J]. Journal of Environmental Radioactivity,2002,62(3):225-233. DOI: 10.1016/S0265-931X(01)00165-5.
[8] Liu Y, Wang Y, Li Y, et al. Earthquake prediction by RBF neural network ensemble[C]//International Symposium on Neural Networks. Berlin, Heidelberg: Springer, 2004: 962-969.10.1007/978-3-540-28648-6_153.
[9] Mirrashid M. Earthquake magnitude prediction by adaptive neuro-fuzzy inference system (ANFIS) based on fuzzy C-means algorithm[J]. Natural Hazards, 2014, 74(3): 1577-1593. DOI: 10.1007/s11069-014-1264-7.
[10] Ikram A, Qamar U. Developing an expert system based on association rules and predicate logic for earthquake prediction[J]. Knowledge-Based Systems, 2015, 75: 87-103. DOI: 10.1016/j.knosys.2014.11.024.
[11] Narayanakumar S, Raja K. A BP artificial neural network model for earthquake magnitude prediction in Himalayas, India[J]. Circuits and Systems, 2016, 7(11): 3456-3468. DOI: 10.4236/cs.2016.711294.
[12] Asim K M, Martínez-Álvarez F, Basit A, et al. Earthquake magnitude prediction in Hindukush region using machine learning techniques[J]. Natural Hazards, 2017, 85(1): 471-486. DOI: 10.1007/s11069-016-2579-3.
[13] Huang J P, Wang X A, Zhao Y, et al. Large earthquake magnitude prediction in Taiwan based on deep learning neural network[J]. Neural Network World, 2018, 28(2): 149-160. DOI: 10.14311/nnw.2018.28.009.
[14] Panakkat A, Adeli H. Recurrent neural network for approximate earthquake time and location prediction using multiple seismicity indicators[J]. Computer-Aided Civil and Infrastructure Engineering, 2009, 24(4): 280-292. DOI: 10.1111/j.1467-8667.2009.00595.x.
[15] Wang Q L, Guo Y F, Yu L X, et al. Earthquake prediction based on spatio-temporal data mining: an LSTM network approach[J]. IEEE Transactions on Emerging Topics in Computing, 2020, 8(1): 148-158. DOI: 10.1109/TETC.2017.2699169.
[16] 李林芳,石耀霖,程术.长短时记忆神经网络在中期地震预报中的探索: 以川滇地区为例[J]. 地球物理学报, 2022, 65(1): 12-25. DOI: 10.6038/cjg2022P0086.
[17] Cheng S, Shi Y. Time series analysis based on machine learning: The cases of sunspot, spring discharge and earthquake prediction[D]. Beijing: University of Chinese Academy of Sciences,2022.
[18] Schorlemmer D, Wiemer S, Wyss M. Variations in earthquake-size distribution across different stress regimes [J]. Nature,2005,437(7058):539-542. DOI: 10.1038/nature04094.
[19] Zamani A, Sorbi M R, Safavi A A. Application of neural network and ANFIS model for earthquake occurrence in Iran[J]. Earth Science Informatics, 2013, 6(2): 71-85. DOI: 10.1007/s12145-013-0112-8.
[20] 董雪情,荆澜涛,田瑞,等.基于LSTM模型的变压器顶层油温预测方法[J].电力学报,2023,38(1):38-45. DOI: 10.13357/j.dlxb.2023.004.
[21] 石耀霖,刘杰,张国民.对我国90年代年度地震预报的评估[J].中国科学院研究生院学报，2000，17(1):63-69.
