본문 바로가기
KERD
논문KCI 등재

온사이트 지진조기경보를 위한 딥러닝 기반 실시간 오탐지 제거

Deep Learning-Based, Real-Time, False-Pick Filter for an Onsite Earthquake Early Warning (EEW) System

이호준|이진구|이우동|이석태|서정범|전인찬|박남률
한국지진공학회|2021.03|한국지진공학회논문집 (Jorunal of the Earthquake Engineering Society of Korea)|제25권 제2호|pp. 71-81|DOI https://doi.org/10.5000/EESK.2021.25.2.071

0

조회

0

다운로드

1

인용

0

참고문헌

키워드

Deep learningConvolutional neural networkFalse alarmEarthquake early warningOnsite warningFalse-pick
주제분류토목공학

초록

This paper presents a real-time, false-pick filter based on deep learning to reduce false alarms of an onsite Earthquake Early Warning (EEW) system. Most onsite EEW systems use P-wave to predict S-wave. Therefore, it is essential to properly distinguish P-waves from noises or other seismic phases to avoid false alarms. To reduce false-picks causing false alarms, this study made the EEWNet Part 1 'False-Pick Filter' model based on Convolutional Neural Network (CNN). Specifically, it modified the Pick_FP (Lomax et al.) to generate input data such as the amplitude, velocity, and displacement of three components from 2 seconds ahead and 2 seconds after the P-wave arrival following one-second time steps. This model extracts log-mel power spectrum features from this input data, then classifies P-waves and others using these features. The dataset consisted of 3,189,583 samples: 81,394 samples from event data (727 events in the Korean Peninsula, 103 teleseismic events, and 1,734 events in Taiwan) and 3,108,189 samples from continuous data (recorded by seismic stations in South Korea for 27 months from 2018 to 2020). This model was trained with 1,826,357 samples through balancing, then tested on continuous data samples of the year 2019, filtering more than 99% of strong false-picks that could trigger false alarms. This model was developed as a module for USGS Earthworm and is written in C language to operate with minimal computing resources.

Abstract

This paper presents a real-time, false-pick filter based on deep learning to reduce false alarms of an onsite Earthquake Early Warning (EEW) system. Most onsite EEW systems use P-wave to predict S-wave. Therefore, it is essential to properly distinguish P-waves from noises or other seismic phases to avoid false alarms. To reduce false-picks causing false alarms, this study made the EEWNet Part 1 'False-Pick Filter' model based on Convolutional Neural Network (CNN). Specifically, it modified the Pick_FP (Lomax et al.) to generate input data such as the amplitude, velocity, and displacement of three components from 2 seconds ahead and 2 seconds after the P-wave arrival following one-second time steps. This model extracts log-mel power spectrum features from this input data, then classifies P-waves and others using these features. The dataset consisted of 3,189,583 samples: 81,394 samples from event data (727 events in the Korean Peninsula, 103 teleseismic events, and 1,734 events in Taiwan) and 3,108,189 samples from continuous data (recorded by seismic stations in South Korea for 27 months from 2018 to 2020). This model was trained with 1,826,357 samples through balancing, then tested on continuous data samples of the year 2019, filtering more than 99% of strong false-picks that could trigger false alarms. This model was developed as a module for USGS Earthworm and is written in C language to operate with minimal computing resources.

참고문헌 (43)

  1. 9.12 Earthquake Response Report, "Jeon YS", Korea Meteorological Administration, 28~ (2017)
  2. Bebbington M, "The linked stress release model for spatiotemporal seismicity: formulations, procedures and applications", Geophysical Journal International, 154(3), pp.925~946 (2003)
  3. Pritchard ME, "New Opportunities to Study Earthquake Precursors", Seismological Research Letters, 91(5), pp.2444~2447 (2020)
  4. Allen MR, "The Potential for Earthquake Early Warning in Southern California", Science, 300(5620), pp.768~789 (2003)
  5. Allen RM, "Earthquake Early Warning: Advances, Scientific Challenges, and Societal Needs", Annual Review of Earth and Planetary Sciences, 47, pp.361~388 (2019)
  6. Allen RM, "Probabilistic warning times for earthquake ground shaking in the San Francisco Bay Area", Seismological Research Letters, 77(3), pp.371~376 (2006)
  7. Hoshiba M, "Outline of the 2011 off the Pacific coast of Tohoku Earthquake (Mw 9.0)—Earthquake Early Warning and observed seismic intensity", Earth, Planets and Space, 63(7) (2011)[DOI]
  8. Cochran ES, "Earthquake early warning ShakeAlertsystem: Testing and certi cation platform", Seismological Research Letters, 89(1), pp.108~117 (2018)
  9. Meier MA, "Reliable Real Time Seismic Signal", Noise Discrimination With Machine Learning, 2019, pp.124, 788~800
  10. Meier MA, "How often can Earthquake Early Warning systems alert sites with high intensity ground motion?", Journal of Geophysical Research: Solid Earth (2020)
  11. Chung A, "Optimizing Earthquake Early Warning Performance: ElarmS‐3", Seismological Research Letters, 90(2A), pp.727~743 (2019)
  12. A Development of Base for Earthquake Early Warning 2nd Step Service (III), "Seo JB", Korea Meteorological Administration, ~ (2018)
  13. Caruso A, "An on-site alert level early warning system for Italy: An onsite early warning system for Italy", Journal of Geophysical Research: Solid Earth (2017)
  14. Wu YM, "Experiment on an Onsite Early Warning Method for the Taiwan Early Warning System", Bulletin of the Seismological Society of America, 95(1), pp.347~353 (2005)
  15. Lomax A, "Automatic Picker Developments and Optimization: FilterPicker-a Robust, Broadband Picker for Real-Time Seismic Monitoring and Earthquake Early Warning", Seismological Research Letters, 83, pp.531~540 (2012)
  16. Vassallo M, "Automatic Picker Developments and Optimization: A Strategy for Improving the Performances of Automatic Phase Pickers", Seismological Research Letters, 83(3), pp.541~554 (2012)
  17. 이호준, "국내 지진관측기록의 P파를 이용한 지진현장경보기술 적용", 한국재난정보학회 논문집, 16(1), pp.133~143 (2020), 한국재난정보학회
  18. Seo JB, "Introduction to the production procedures and technologies of KMA intensity information", Korea Meteorological Administration, 56~59 (2019)
  19. Samuel AL, "Some studies in machine learning using the game of checkers", IBM Journal of Research and Development, 3(3), pp.210~229 (1959)
  20. Perol T, "Convolutional neural network for earthquake detection and location", Science Advances, 4(2) (2018)
  21. Wang B, "Deep-learning-based seismic data interpolation: A preliminary result", GEOPHYSICS, 84(1) (2019)
  22. Derakhshani A, "Predicting the principal strong ground motion parameters: A deep learning approach", Applied Soft Computing, 80, pp.192~201 (2019)
  23. Seydoux L, "Clustering earthquake signals and background noises in continuous seismic data with unsupervised deep learning", Nat Commun, 11, pp.3972~ (2020)
  24. Araya-Polo M, "Deep-learning tomography", The Leading Edge, 37(1), pp.2~80 (2018)
  25. Kong Q, "2018 Seismology of the Americas Meeting:Technical Sessions:Applications of Machine Learning and Data Science in Seismology", Seismological Research Letters, 89(2B), pp.728~729 (2018)
  26. Hsu TY, "Performance of the NCREE’s on‐site warning system during the 5 February 2016Mw 6.53 Meinong earthquake", Geophysical Research Letters, 43(17) (2016)
  27. "Earthquake prediction and warning in 5 seconds by artificial intelligence... AI earthquake early warning system released in Taiwan" (2018)[URL]
  28. Ross ZE, "Generalized Seismic Phase Detection with Deep Learning", Bulletin of the Seismological Society of America, 108(5A), pp.2894~2901 (2019)
  29. Ronneberger 0, "U-Net: Convolutional Networks for Biomedical Image Segmentation", Miccai, 234~241 (2015)
  30. "2018", ~
  31. 김승일, "다중 주파수 대역 convolutional neural network 기반지진 신호 검출 기법", 한국음향학회지, 38(1), pp.23~29 (2019), 한국음향학회[DOI]
  32. Géron A, "Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow" (2019), O'Reilly Media, Inc
  33. Seo JB, "Development of Onsite warning technology using Deep Learning", Proceedings of EESK Conference 2020 (2020)
  34. Kennett BLN, "Travel times for global earthquake location and phase association", Geophysical Journal International, 105, pp.429~465 (1991)
  35. Abadi M, "TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems", ~ (2016)
  36. Krizhevsky A, "Imagenet classification with deep convolutional neural networks", Advances in Neural Information Processing Systems 25 (NIPS 2012), pp.1097~1105 (2012)
  37. Allen RV, "Automatic earthquake recognition and timing from single traces", Bulletin of the Seismological Society of America, 68(5), pp.1521~1532 (1978)
  38. Nair V, "Rectified linear units improve restricted boltzmann machines", 27th International Conference on Machine Learning (ICML 2010) (2010)
  39. Srivastava N, "Improving neural networks with dropout", Ph.D, pp.2013
  40. Goodfellow I, "Deep Learning" (2016), MIT Press
  41. Fernández-Redondo M, "Weight initialization methods for multilayer feedforward", ESANN, pp.119~124 (2001)
  42. Seo JB, "Method for determining real-time alarm for local target earthquake early warning using mechanical learning", Korean patent 10-2020-0147842, ~
  43. "Volcano, Earthquake, Tsunami"[URL]

관련 논문AI추천

지진 이벤트 분류를 위한 Bottleneck 구조의 심층 CNN 기법고한석, 구본화, 김관태, 민정기한국컴퓨터정보학회 · 한국컴퓨터정보학회논문지2019 · 제24권 제1호특성 자료를 활용한 한반도 동남권 지역의 최대지반가속도 예측 연구박정호, 유성화, 윤여웅, 임인섭한국지진공학회 · 한국지진공학회논문집2025 · 제29권 제2호진원지 인근 지진 조기 경보를 위한 선착 P파 다중 탐지 시스템 개발김영석, 노진석, 이태희, 홍승서한국지반신소재학회 · 한국지반신소재학회 논문집2019 · 제18권 제4호Determination of High-pass Filter Frequency with Deep Learning for Ground Motion서정범, 이진구, 전성진한국지진공학회 · 한국지진공학회논문집2024 · 제28권 제4호근거리 지진관측자료의 S파를 이용한 지진규모 평가 연구연관희, 이강렬, 최신규한국지진공학회 · 한국지진공학회논문집2024 · 제28권 제3호잡음 자료 취득을 통한 기계학습 기반 현장 탄성파 자료 잡음 억제 연구김충호, 김한준, 전형구한국자원공학회 · 한국자원공학회지2021 · 제58권 제5호

이 간행물의 다른 논문 (3건)

지진 손상 상관성이 플랜트의 확률론적 지진 안전성 평가에 미치는 영향곽신영, 임승현, 박동욱, 최인길, 전법규pp. 53-58지진동 모사를 통한 역사지진 규모와 진앙 평가김성룡, 이상준pp. 59-69일본 원전 내진설계 기술기준을 적용한 모의지진파(가속도시간이력) 작성문일환, 진승민, 이용선, 김용복pp. 83-91
원문 보기