Discrimination of earthquake and microtremor based on generative adversarial network and random forest combination model
LIU Heyi1,2, SONG Jindong1,2, LI Shanyou1,2
1.Key Laboratory of Earthquake Engineering and Engineering Vibration, Institute of Engineering Mechanics, China Earthquake Administration, Harbin 150080, China;
2.Key Laboratory of Earthquake Disaster Mitigation, Ministry of Emergency Management, Harbin 150080, China
摘要地震事件识别是地震监测业务的基础性工作，特别是随着大规模地震台站建设、海量地震数据汇聚以及地震预警的发展，从连续波形记录中自动区分地震与地脉动噪声显得更加重要。本文以准确识别地震事件为目标，提出了一种基于生成对抗网络与随机森林的地震事件识别组合模型，该模型先利用生成对抗网络提取波形信号特征、再利用随机森林基于提取的波形信号特征将地震事件识别转化为地震与地脉动的分类问题。地震与地脉动各5378条数据的测试集研究结果表明，该模型对地震事件与地脉动的分类准确率均可以达到99%以上，地震事件识别率比较传统的长短时窗方法（STA/LTA，short term averaging/long term averaging）提高了23.56个百分点，表明该模型可以从地脉动中准确识别地震事件，并在地震监测与地震预警中具有应用前景。
关键词：生成对抗网络；随机森林；地震；地脉动；地震预警
Abstract：Seismic event discrimination is one of the basic tasks of earthquake monitoring. Especially with the construction of large-scale seismic stations, the accumulation of massive seismic data and the development of earthquake early warning (EEW), it is more important to distinguish earthquakes accurately and automatically from continuous waveform records. To accurately discriminating seismic events, this paper proposes a combined model for seismic event discrimination based on a generative adversarial network (GAN) and a Random Forest (RF). The model first uses a GAN to extract the waveform features, and then uses a random forest to transform the seismic event discrimination into the classification problem of earthquakes and microtremor based on the extracted waveform features. The results of the testing set which has 5378 data of earthquakes and microtremor show that, the classification accuracy of the combined model for classifying seismic events and microtremor can reach more than 99%, and the results has increased by 23.56% compared with the STA/LTA (short term averaging/long term averaging) algorithm, which proves that the model can accurately identify seismic events from microtremor and has application prospects in earthquake monitoring and EEW.
Key words: Generative adversarial network; Random Forest; Earthquake; Microtremor; Earthquake early warning
刘赫奕1,2，宋晋东1,2，李山有1,2. 基于生成对抗网络与随机森林组合模型的地震与地脉动区分研究[J]. 振动与冲击, 2022, 41(15): 312-324.
LIU Heyi1,2, SONG Jindong1,2, LI Shanyou1,2. Discrimination of earthquake and microtremor based on generative adversarial network and random forest combination model. JOURNAL OF VIBRATION AND SHOCK, 2022, 41(15): 312-324.
[1] Allen R V. Automatic earthquake recognition and timing from single traces [J]. Bulletin of the Seismological Society of America, 1978, 68(5): 1521-1532.
[2] Sleeman R, and Van Eck T. Robust automatic P-phase picking: an on-line implementation in the analysis of broadband seismogram recordings [J]. Physics of the earth planetary interiors, 1999, 113(1-4): 265-275.
[3] Saragiotis C D, Hadjileontiadis, L J, and Panas S M. PAI-S/K: A robust automatic seismic P phase arrival identification scheme [J]. IEEE Transactions on Geoscience Remote Sensing, 2002, 40(6): 1395-1404.
[4] Brown J R, Beroza G C, and Shelly D R. An autocorrelation method to detect low frequency earthquakes within tremor [J]. Geophysical Research Letters, 2008, 35(16).
[5] Peng Z, Zhao P. Migration of early aftershocks following the 2004 Parkfield earthquake [J]. Nature Geoence, 2009, 2(12):877-881.
[6] Zefeng Li, Men-Andrin Meier, Egill Hauksson, et al. Machine Learning Seismic Wave Discrimination: Application to Earthquake Early Warning [J]. Geophysical Research Letters, 2018, 45(10).
[7] 赵明, 陈石.基于深度学习卷积神经网络的地震波形自动分类与识别[J].地球物理学报,2019,62(1): 374-382.
ZHAO ming, CHEH shi. Waveform classification and seismic recognition by convolution neural network [J]. Chinese Journal of Geophysics, 2019, 62(1):374-382. (in Chinese)
[8] Mousavi S M, Ellsworth W L, Zhu W, et al. Earthquake transformer—an attentive deep-learning model for simultaneous earthquake detection and phase picking [J]. Nature communications, 2020, 11(1): 1-12.
[9] Kong Q, Allen R M, Schreier L, et al. MyShake: A smartphone seismic network for earthquake early warning and beyond [J]. Science Advances, 2016, 2(2): e1501055 - e1501055.
[10] 江汶乡,于海英,赵晓芬,等.用于地震预警系统的单台站防误触发算法研究[J].自然灾害学报,2015,24(02):23-31.
JIANG wenxiang, YU haiying, ZHAO xiaofen, et al. Interference Prevention Algorithm Used for Single Station of Earthquake Early Warning System [J]. Journal of Natural Disasters, 2015, 24(02): 23-31. (in Chinese)
[11] Perol, T., Gharbi, M., Denolle, M. Convolutional neural network for earthquake detection and location [J]. Science advances, 2018, 4(2): e1700578.
[12] Wu Yue, Lin Youzuo, Zhou Zheng, et al. DeepDetect: A cascaded region-based densely connected network for seismic event detection [J]. IEEE Transactions on Geoscience and Remote Sensing, 2018, 57(1): 62-75.
[13] 马强,金星,李山有,等.用于地震预警的P波震相到时自动拾取[J].地球物理学报,2013,56(07):2313-2321.
MA Qiang, JIN Xing, LI Shanyou, et al. Automatic P-arrival Detection for Earthquake Early Warning [J]. Chinese Journal of Geophysics, 2013, 56(7): 2313-2321. (in Chinese)
[14] Goodfellow I J, Pouget-Abadie J, Mirza M, et al. Generative Adversarial Nets [C]// International Conference on Neural Information Processing Systems. 2014.
[15] Arjovsky, M., Chintala, S., Bottou, L. Wasserstein Generative Adversarial Networks [C]// Proceedings of the 34th International Conference on Machine Learning. PMLR, 2017, 70, 214–223.
[16] 周志华.机器学习[M].清华大学出版社,2016,179-181.
ZHOU Zhihua. Machine Learning [M]. Tsinghua University Press, 2016, 179-181. (in Chinese)
[17] Tin Kam Ho. Random decision forests[C]// Document Analysis and Recognition, 1995. Proceedings of the Third International Conference on. IEEE Computer Society, 1995.
[18] Ho T K. The Random Subspace Method for Constructing Decision Forests [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1998, 20(8):0-844.
[19] Creswell A, White T, Dumoulin V, et al. Generative Adversarial Networks: An Overview [J]. IEEE Signal Processing Magazine, 2017, 35(1):53-65.
[20] Swami A, Jain R. Scikit-learn: Machine Learning in Python [J]. Journal of Machine Learning Research, 2012, 12(10):2825-2830.)
[21] 何先龙, 佘天莉, 高峰.一种地震 P 波和 S 波初至时间自动拾取的新方法[J].地球物理学报,2016,59(7): 2519-2527.
HE xianlong, SHE tianli, GAO feng. A new method for picking up arrival times of seismic P and S waves automatically[J]. Chinese Journal of Geophysics, 2016, 59(7): 2519-252. (in Chinese)
