Machine learning prediction models for ground motion parameters and seismic damage assessment of buildings at a regional scale

Abstract

This study examines the feasibility of using a machine learning approach for rapid damage assessment of rein-forced concrete (RC) buildings after the earthquake. Since the real-world damaged datasets are lacking, have limited access, or are imbalanced, a simulation dataset is prepared by conducting a nonlinear time history analy-sis. Different machine learning (ML) models are trained considering the structural parameters and ground motion characteristics to predict the RC building damage into five categories: null, slight, moderate, heavy, and collapse. The random forest classifier (RFC) has achieved a higher prediction accuracy on testing and real-world damaged datasets. The structural parameters can be extracted using different means such as Google Earth, Open Street Map, unmanned aerial vehicles, etc. However, recording the ground motion at a closer distance requires the installation of a dense array of sensors which requires a higher cost. For places with no earthquake recording station/device, it is difficult to have ground motion characteristics. For that different ML-based regressor models are developed utilizing past-earthquake information to predict ground motion parameters such as peak ground acceleration and peak ground velocity. The random forest regressor (RFR) achieved better results than other regression models on testing and validation datasets. Furthermore, compared with the results of similar research works, a better result is obtained using RFC and RFR on validation datasets. In the end, these models are uti-lized to predict the damage categories of RC buildings at Saitama University and Okubo Danchi, Saitama, Japan after an earthquake. This damage information is crucial for government agencies or decision-makers to respond systematically in post-disaster situations.

Keywords

Seismic damage prediction / Ground motion parameter / Machine learning algorithms / Nonlinear time history analysis / RC buildings

References

[1] UN News, 21 Feb 2020. URL: news. un. org/en/story/2023/02/1133717 Accessed: 01 August 2022.

[2] Internet geography, Nepal earthquake 2015. URL: www. internetgeography.net/topics/nepal-earthquake-2015/#:-:text=The%20primary%20effects%20of%20the%202015%20earthquake%20in%20Nepal%20include,along%20with%20communications%2C%20were%20 affected Accessed: 01 August 2022

[3] Riedel I, Guéguen P. Modeling of damage-related earthquake losses in a moderate seismic-prone country and cost-benefit evaluation of retrofit investments: applica-tion to France. Nat Hazard 2018;90:639-62.

[4] Sajan KC, Bhusal A, Gautam D, Rupakhety R. Earthquake damage and rehabilitation intervention prediction using machine learning. Eng Fail Anal 2023;144:106949.

[5] Hoskere V., Narazaki Y., Hoang T.A. and Spencer Jr, B.F., 2018. Towards auto-mated post-earthquake inspections with deep learning-based condition-aware mod-els. arXiv preprint arXiv:1809.09195.

[6] Applied Technology Council ATC., 1985. Earthquake damage evaluation data for California (ATC-13), Redwood, CA

[7] De Luca F, Verderame GM, Manfredi G.Analytical versus observational fragili-ties: the case of Pettino (L’Aquila) damage data database. Bull Earthq Eng 2015;13:1161-81.

[8] Rota M, Penna A, Magenes G. A methodology for deriving analytical fragility curves for masonry buildings based on stochastic nonlinear analyses. Eng Struct 2010;32(5):1312-23.

[9] Domaneschi M, Noori AZ, Pietropinto MV, Cimellaro GP. Seismic vulnerability as-sessment of existing school buildings. Comput Struct 2021;248:106522.

[10] Mangalathu S. Performance based grouping and fragility analysis of box-girder bridges in California. Georg Instit Technol 2017;201(7).

[11] Calvi GM, Pinho R, Magenes G, Bommer JJ, Restrepo-Vélez LF, Crowley H.Devel-opment of seismic vulnerability assessment methodologies over the past 30 years. ISET J Earthq Technol 2006;43( 3):75-104.

[12] Hansapinyo C, Latcharote P, Limkatanyu S. Seismic building damage prediction from GIS-based building data using artificial intelligence system. Front Built Env-iron 2020;6:576919.

[13] Baker JW, Allin Cornell C. A vector-valued ground motion intensity mea-sure consisting of spectral acceleration and epsilon. Earthq Eng Struct Dyn 2005;34(10):1193-217.

[14] Castellazzi G, Gentilini C, Nobile L. Seismic vulnerability assessment of a histor-ical church: limit analysis and nonlinear finite element analysis. Adv Civil Eng 2013;2013.

[15] Preciado A. Seismic vulnerability and failure modes simulation of ancient masonry towers by validated virtual finite element models. Eng Fail Anal 2015;57:72-87.

[16] Castori G, Borri A, De Maria A, Corradi M, Sisti R. Seismic vulnerability assessment of a monumental masonry building. Eng Struct 2017;136:454-65.

[17] Ahmed A, Shahzada K. October. Seismic vulnerability assessment of confined ma-sonry structures by macro-modeling approach. Structures (Vol. 27, pp. 639-649). Elsevier; 2020.

[18] Ceroni F, Pecce M, Sica S, Garofano A. Assessment of seismic vulnerability of a historical masonry building. Buildings 2012;2(3):332-58.

[19] Ahmed B, Mangalathu S, Jeon JS. Seismic damage state predictions of reinforced concrete structures using stacked long short-term memory neural networks. J Build Eng 2022;46:103737.

[20] Silva V, Akkar S, Baker J, Bazzurro P, Castro JM, Crowley H, Dolsek M, Galasso C, Lagomarsino S, Monteiro R, Perrone D. Current challenges and future trends in an-alytical fragility and vulnerability modeling. Earthq Spectra 2019;35(4):1927-52.

[21] Harirchian E, Hosseini SEA, Jadhav K, Kumari V, Rasulzade S, I şı k E, Wasif M, Lahmer T. A review on application of soft computing techniques for the rapid vi-sual safety evaluation and damage classification of existing buildings. J Build Eng 2021;43:102536.

[22] Xie Y, Ebad Sichani M, Padgett JE, DesRoches R. The promise of implementing ma-chine learning in earthquake engineering: a state-of-the-art review. Earthq Spectra 2020;36(4):1769-801.

[23] Sun H, Burton HV, Huang H. Machine learning applications for building struc-tural design and performance assessment: state-of-the-art review. J Build Eng 2021;33:101816.

[24] Zhang H, Cheng X, Li Y, He D, Du X. Rapid seismic damage state assessment of RC frames using machine learning methods. J Build Eng 2023;65:105797.

[25] Kostinakis K, Morfidis K, Demertzis K, Iliadis L. Classification of buildings’ poten-tial for seismic damage using a machine learning model with auto hyperparameter tuning. Eng Struct 2023;290:116359.

[26] Bhatta S, Dang J. Seismic damage prediction of RC buildings using machine learning. Earthq Eng Struct Dyn 2023.

[27] Bhatta S, Dang J. Machine learning-based classification for rapid seismic damage assessment of buildings at a regional scale. J Earthq Eng 2023:1-31.

[28] Bhatta S, Dang J. Multiclass seismic damage detection of buildings using quantum convolutional neural network. Comput-Aided Civ Infrastruct Eng 2023.

[29] Xu Y, Li Y, Zheng X, Zheng X, Zhang Q. Computer-vision and machine-learn-ing-based seismic damage assessment of reinforced concrete structures. Buildings 2023;13(5):1258.

[30] Tang Q, Dang J, Cui Y, Wang X, Jia J. Machine learning-based fast seismic risk assessment of building structures. J Earthq Eng 2022;26(15):8041-62.

[31] Molas GL, Yamazaki F. Neural networks for quick earthquake damage estimation. Earthq Eng Struct Dyn 1995;24(4):505-16.

[32] Kiani J, Camp C, Pezeshk S. On the application of machine learning techniques to derive seismic fragility curves. Comput Struct 2019;218:108-22.

[33] Xu Y, Lu X, Tian Y, Huang Y. Real-time seismic damage prediction and comparison of various ground motion intensity measures based on machine learning. J Earthq Eng 2022;26(8):4259-79.

[34] Arslan MH. An evaluation of effective design parameters on earthquake performance of RC buildings using neural networks. Eng Struct 2010;32(7):1888-98.

[35] Arslan MH, Ceylan M, Koyuncu T. Determining earthquake performances of ex-isting reinforced concrete buildings by using ANN. Int J Civil Environ Eng 2015;9(8):1097-101.

[36] Zhang Y, Burton HV, Sun H, Shokrabadi M. A machine learning framework for as-sessing post-earthquake structural safety. Struct Saf 2018;72:1-16.

[37] Roeslin S, Ma Q, Juárez-Garcia H, Gómez-Bernal A, Wicker J, Wotherspoon L.A machine learning damage prediction model for the 2017 Puebla-Morelos, Mexico, earthquake. Earthq Spectra 2020;36(2_suppl):314-39.

[38] Mangalathu S, Sun H, Nweke CC, Yi Z, Burton HV. Classifying earthquake damage to buildings using machine learning. Earthq Spectra 2020;36(1):183-208.

[39] Stojadinovi ćZ, Kova čevi ćM, Marinkovi ćD, Stojadinovi ćB. Rapid earthquake loss assessment based on machine learning and representative sampling. Earthq Spectra 2022;38(1):152-77.

[40] Kermani E., Jafarian Y. and Baziar, M.H., 2009. New predictive models for the vmax/amax ratio of strong ground motions using genetic programming.

[41] Ghimire S, Guéguen P, Giffard-Roisin S, Schorlemmer D.Testing machine learn-ing models for seismic damage prediction at a regional scale using building-dam-age dataset compiled after the 2015 Gorkha Nepal earthquake. Earthq Spectra 2022;38(4):2970-93.

[42] Government of Japan Japan’s Natural Disaster Early Warning Systems, and Inter-national Cooperative Efforts. Early Warning Sub-Committee of the Inter-Ministerial Committee on International Cooperation for Disaster Reduction; 2006.

[43] Boore DM, Atkinson GM. Stochastic prediction of ground motion and spectral re-sponse parameters at hard-rock sites in eastern North America. Bull Seismol Soc Am 1987;77(2):440-67.

[44] Midorikawa S. Preliminary analysis for attenuation of peak ground velocity on stiff site. Proc of Int Worksh Strong Motion Data 1993;2:39-48.

[45] Joshi A, Midorikawa S. Attenuation characteristics of ground motion intensity from earthquakes with intermediate depth. J Seismol 2005;9:23-37.

[46] Joshi A, Raman B, Mohan CK, Cenkeramaddi LR. Application of a new machine learning model to improve earthquake ground motion predictions. Nat Hazards 2023:1-25.

[47] Boore DM, Joyner WB. The empirical prediction of ground motion. Bull Seismol Soc Am 1982;72(6B):S43-60.

[48] Atkinson GM. Empirical attenuation of ground-motion spectral amplitudes in south-eastern Canada and the north-eastern United States. Bull Seismol Soc Am 2004;94(3):1079-95.

[49] Zhang B, Yu Y, Li X, Wang Y. Ground motion prediction equation for the average horizontal component of PGA, PGV, and 5% damped acceleration response spectra at periods ranging from 0.033 to 8.0 s in southwest China. Soil Dyn Earthq Eng 2022;159:107297.

[50] Kumar A, Ghosh G, Gupta PK, Kumar V, Paramasivam P. Seismic hazard anal-ysis of Silchar city located in North East India. Geomatics, Nat Hazards Risk 2023;14(1):2170831.

[51] Khosravikia F, Clayton P. Machine learning in ground motion prediction. Comput Geosci 2021;148(January):104700. doi: 10.1016/j.cageo.2021.104700.

[52] Kubo H, Kunugi T, Suzuki W, Suzuki S, Aoi S. Hybrid predictor for ground-motion in-tensity with machine learning and conventional ground motion prediction equation. Sci Rep 2020;10(1):11871.

[53] Khosravikia F, Zeinali Y, Nagy Z, Clayton P, Rathje EM. Neural network-based equa-tions for predicting PGA and PGV in Texas, Oklahoma, and Kansas. In: Geotechnical earthquake engineering and soil dynamics v. Reston, VA: American Society of Civil Engineers; 2018. p. 538-49.

[54] Shoushtari AV, Adnan AB, Zare M. Ground motion prediction equations for distant subduction interface earthquakes based on empirical data in the Malay Peninsula and Japan. Soil Dyn Earthq Eng 2018;109:339-53.

[55] Wang A, Li S, Lu J, Zhang H, Wang B, Xie Z. Prediction of PGA in earthquake early warning using a long short-term memory neural network. Geophys J Int 2023;234(1):12-24.

[56] Kim S, Hwang Y, Seo H, Kim B. Ground motion amplification models for Japan using machine learning techniques. Soil Dyn Earthq Eng 2020;132:106095.

[57] Ohno S, Tsuruta R. Ground-motion prediction by ANN using machine learning for the Tohoku region, Japan. In:11th National conference on earthquake engineering 2018: integrating science, engineering, and policy, NCEE 2018. Earthquake Engi-neering Research Institute; 2018. p. 5429-37.

[58] Zhu J, Li S, Song J. Hybrid deep-learning network for rapid on-site peak ground velocity prediction. IEEE Trans Geosci Remote Sens 2022;60:1-12.

[59] Goda K, Kiyota T, Pokhrel RM, Chiaro G, Katagiri T, Sharma K, Wilkinson S. The 2015 Gorkha Nepal earthquake: insights from earthquake damage survey. Front Built Environ 2015; 1:8.

[60] Chelidze T, Kiria T, Melikadze G, Jimsheladze T, Kobzev G. Earthquake forecast as a machine learning problem for imbalanced datasets: example of Georgia, Caucasus. Front Earth Sci 2022;10:847808.

[61] Lu X, Han B, Hori M, Xiong C, Xu Z. A coarse-grained parallel approach for seismic damage simulations of urban areas based on refined models and GPU/CPU cooper-ative computing. Adv Eng Softw 2014;70:90-103.

[62] Maharjan N, Dang J, Shrestha A.Machine Learning Based Structural Seismic Damage Detection From Monitoring Acceleration Data. 17th World Conference on Earthquake Engineering, 17WCEE, Sendai, Japan; 2020.

[63] NIED. The NIED Strong-Motion Seismograph Networks. 2020. http://www.kyoshin.bosai.go.jp. Accessed March 1, 2020.

[64] Kanno T, Narita A, Morikawa N, Fujiwara H, Fukushima Y. A new attenuation re-lation for strong ground motion in Japan based on recorded data. Bull Seismol Soc Am 2006;96(3):879-97.

[65] Han SW, Kim WT, Foutch DA. Tensile strength equation for HSS bracing members having slotted end connections. Earthq Eng Struct Dyn 2007;36(8):995-1008.

[66] Vamvatsikos D, Cornell CA. Incremental dynamic analysis. Earthq Eng Struct Dyn 2002;31(3):491-514.

[67] Bommer JJ, Acevedo AB. The use of real earthquake accelerograms as input to dy-namic analysis. J Earthq Eng 2004;8(spec01):43-91.

[68] Watson-Lamprey J, Abrahamson N. Selection of ground motion time series and limits on scaling. Soil Dyn Earthq Eng 2006;26(5):477-82.

[69] Luco N, Bazzurro P. Does amplitude scaling of ground motion records result in biased nonlinear structural drift responses? Earthq Eng Struct Dyn 2007;36(13):1813-35.

[70] Gunturi SKV. Building specific damage estimation. Proc 10WCEE 1992;10:6001-6.

[71] Multi-hazard loss estimation methodology: earthquake modelHAZUS technical manual. FEMA; 2003.

[72] Hastie T, Tibshirani R, Friedman J. The elements of statistical learning. New York, NY, USA: Springer series in statistics; 2001.

[73] Breiman L. ST4_Method_Random_Forest. Mach Learn 2001;45(1):5-32.

[74] Fawagreh K, Gaber MM, Elyan E. Random forests: from early developments to recent advancements. Syst Sci Control Eng: Open Access J 2014;2(1):602-9.

[75] Breiman L. Bagging predictors. Mach Learn 1996;24:123-40.

[76] Breiman L. Random forests. Mach Learn 2001;45:5-32.

[77] Grünthal G.,1998 European macroseismic scale 1998 (EMS-98).

[78] United States Geological Survey USGS.https://earthquake.usgs.gov/earthquakes/eventpage/us20002926/shakemap/intensity. Accessed April 5, 2022.

[79] Lin X, Chen X, Wang F, He X, Zhang L, Skalomenos KA.Real-time physics-based regional earthquake disaster simulation using structural monitoring data. 17th World Conference on Earthquake Engineering (17WCEE) Sendai, Japan; 2020.

[80] American Society of Civil Engineers. Minimum design loads for buildings and other structures ASCE7-10

[81] Fukushima Y. Revised attenuation relation for peak horizontal acceleration using a new data base. Program Abstr Seism Soc Jpn 1992;116.

[82] Sokolov V, Wenzel F. Further analysis of the influence of site conditions and earth-quake magnitude on ground-motion within-earthquake correlation: analysis of PGA and PGV data from the K-NET and the KiK-net (Japan) networks. Bull Earthq Eng 2013;11:1909-26.

[83] Pokhrel RM, Kuwano J, Tachibana S. Liquefaction hazard zonation mapping of the Saitama City, Japan. J Nepal Geol Soc 2010;40:69-76.