Earthquakes and Structures

  • 제25권6호
  • /
  • Pages.469-479
  • /
  • 2023
  • /
  • 2092-7614(pISSN)
  • /
  • 2092-7622(eISSN)
테크노프레스 (Techno-Press)
권호 표지
DOI QR코드

DOI QR Code

Utilization of deep learning-based metamodel for probabilistic seismic damage analysis of railway bridges considering the geometric variation

  • Xi Song (Department of Civil and Architectural Engineering & Construction Management, Milwaukee School of Engineering) ;
  • Chunhee Cho (Department of Civil and Environmental Engineering, University of Hawaii at Manoa) ;
  • Joonam Park (Department of Civil and Environmental Engineering, Wonkwang University)
  • 투고 : 2022.07.07
  • 심사 : 2023.11.09
  • 발행 : 2023.12.25
⟨ 이전 논문 다음 논문 ⟩

초록

A probabilistic seismic damage analysis is an essential procedure to identify seismically vulnerable structures, prioritize the seismic retrofit, and ultimately minimize the overall seismic risk. To assess the seismic risk of multiple structures within a region, a large number of nonlinear time-history structural analyses must be conducted and studied. As a result, each assessment requires high computing resources. To overcome this limitation, we explore a deep learning-based metamodel to enable the prediction of the mean and the standard deviation of the seismic damage distribution of track-on steel-plate girder railway bridges in Korea considering the geometric variation. For machine learning training, nonlinear dynamic time-history analyses are performed to generate 800 high-fidelity datasets on the seismic response. Through intensive trial and error, the study is concentrated on developing an optimal machine learning architecture with the pre-identified variables of the physical configuration of the bridge. Additionally, the prediction performance of the proposed method is compared with a previous, well-defined, response surface model. Finally, the statistical testing results indicate that the overall performance of the deep-learning model is improved compared to the response surface model, as its errors are reduced by as much as 61%. In conclusion, the model proposed in this study can be effectively deployed for the seismic fragility and risk assessment of a region with a large number of structures.

키워드

과제정보

This research was funded by Wonkwang University in 2021.

참고문헌

  1. Alam, J. Kim, D. and Choi, B. (2017), "Seismic probabilistic risk assessment of weir structures considering the earthquake hazard in the Korean Peninsula", Earthq. Struct., 13(4), 421-427. https://doi.org/10.12989/eas.2017.13.4.421.
  2. Brownlee, J. (2019), A Gentle Introduction to Dropout for Regularizing Deep Neural Networks; Machine Learning Mastery, San Juan, Puerto Rico. https://machinelearningmastery.com/dropout-for-regularizing-deep-neural-networks
  3. Buratti, N., Ferracuti, B. and Savoia, M. (2010), "Response surface with random factors for seismic fragility of reinforced concrete frames", Struct. Saf., 32(1), 42-51. https://doi.10.1016/j.strusafe.2009.06.003.
  4. Cao, A.T., Nahar, T.T., Kim, D. and Choi, B. (2019), "Earthquake risk assessment of concrete gravity dam by cumulative absolute velocity and response surface methodology", Earthq. Struct., 17(5), 511-519. https://doi.org/10.12989/eas.2019.17.5.511.
  5. Dertat, A. (2018), Applied Deep Learning - Part 1: Artificial Neural Networks; Medium, San Francisco, CA, USA. https://towardsdatascience.com/applied-deep-learning-part-1-artificial-neural-networks-d7834f67a4f6
  6. Ellingwood, B.R. (2001), "Earthquake risk assessment of building structures", Reliab. Eng. Syst. Saf., 74(3), 251-262. https://doi.org/10.1016/S0951-8320(01)00105-3.
  7. Franchin, P., Lupoi, A., Pinto, P.E. and Schotanus, M.I. (2003), "Seismic fragility of reinforced concrete structures using a response surface approach", J. Earthq. Eng., 7(1), 45-77. https://doi.10.1080/13632460309350473.
  8. Gehl, P., Seyedi, D.M. and Douglas, J. (2013), "Vector-valued fragility functions for seismic risk evaluation", Bull. Earthq. Eng., 11(2), 365-384. https://doi.10.1007/s10518-012-9402-7.
  9. Goodfellow, I., Bengio, Y. and Courville, A. (2016), Deep Learning (Adaptive Computation and Machine Learning Series), Illustrated Edition, The MIT Press, Cambridge, MA, USA.
  10. Han, S.W. and Choi, Y.S, (2008), "Seismic hazard analysis in low and moderate seismic region - Korean peninsula", Struct. Saf., 30(6), 543-558. https://doi.org/10.1016/j.strusafe.2007.10.004.
  11. Hu, Y., Zhao, J., Zhang, D. and Zhang, Y. (2018), "Seismic risk assessment of concrete-filled double-skin steel tube/moment-resisting frames", Earthq. Struct., 14(3), 249-259. https://doi.org/10.12989/eas.2018.14.3.249.
  12. Huang, C. and Huang, S. (2020), "Predicting capacity model and seismic fragility estimation for RC bridge based on artificial neural network", Struct., 27, 1930-1939. https://doi.org/10.1016/j.istruc.2020.07.063.
  13. Iman, R.L. and Conover, W.J. (1980), "Small sample sensitivity analysis techniques for computer models with an application to risk assessment", Commun. Theory Method., 9(17), 1749-1842. https://doi.org/10.1080/03610928008827996.
  14. Krizhevsky, A., Sutskever, I. and Hinton, G.E. (2017), "ImageNet classification with deep convolutional neural networks", Commun. ACM, 60(6), 84-90. https://doi.org/10.1145/3065386.
  15. Korea Railroad Research Institute (2008), "Seismic performance evaluation of Korean railway infrastructures", Technical Report, Korea Railroad Research Institute.
  16. Mangalathu, S., Heo, G. and Jeon, J.S. (2018), "Artificial neural network based multi-dimensional fragility development of skewed concrete bridge classes", Eng. Struct., 162, 166-176. https://doi.org/10.1016/j.engstruct.2018.01.053.
  17. Mangalathu, S. and Jeon, J. (2019), "Stripe-based fragility analysis of multispan concrete bridge classes using machine learning techniques", Earthq. Eng. Struct. Dyn., 48(11), 1238-1255. https://doi.org/10.1002/eqe.3183.
  18. Mazzoni, S., McKenna, F., Scott, M.H. and Fenves, G.L. (2007), OpenSees Command Language Manual, University of California, Berkeley, Berkeley, CA, USA.
  19. Nielson, B.G. and DesRoches, R. (2007), "Analytical seismic fragility curves for typical bridges in the central and southeastern United States", Earthq. Spectra, 23(3), 615-633. https://doi.org/10.1193/1.2756815.
  20. Padgett, J.E., Ghosh, J. and Duenas-Osorio, L. (2013), "Effects of liquefiable soil and bridge modelling parameters on the seismic reliability of critical structural components", Struct. Infrastr. Eng., 9(1), 1-19. https://doi.org/10.1080/15732479.2010.524654.
  21. Pang, Y., Dang, X. and Yuan, W. (2014), "An artificial neural network based method for seismic fragility analysis of highway bridges", Adv. Struct. Eng., 17(3), 413-428. https://doi.org/10.1260/1369-4332.17.3.413.
  22. Polson, N. and Sokolov, V. (2020), "Deep learning: Computational aspects", WIREs Comput. Stat., 12(5), e1500. https://doi.org/10.1002/wics.1500.
  23. Park, J. and Choi, E. (2011), "Fragility analysis of track-on steelplate-girder railway bridges in Korea", Eng. Struct., 33(3), 696-705. https://doi.org/10.1016/j.engstruct.2010.09.028.
  24. Park, J., Towashiraporn, P., Craig, J.I. and Goodno, B.J. (2009), "Seismic fragility analysis of low-rise unreinforced masonry structures", Eng. Struct., 31(1), 125-137. https://doi.org/10.1016/j.engstruct.2008.07.021.
  25. Park, J. and Towashiraporn, P. (2014), "Rapid seismic damage assessment of railway bridges using the response-surface statistical model", Struct. Saf., 47(2), 1-12. https://doi.org/10.1016/j.strusafe.2013.10.001.
  26. Sainct, R., Feau, C., Martinez, J.M. and Garnier, J. (2020), "Efficient methodology for seismic fragility curves estimation by active learning on support vector machines", Struct. Saf., 86, 101972. https://doi.org/10.1016/j.strusafe.2020.101972.
  27. Seo, J. and Linzell, D.G. (2013), "Use of response surface metamodels to generate system level fragilities for existing curved steel bridges", Eng. Struct., 52, 642-653. https://doi.org/10.1016/j.engstruct.2013.03.023.
  28. Singhal, A. and Kiremidjian, A.S. (1996), "Method for probabilistic evaluation of seismic structural damage", J. Struct. Eng., 122(12), 1459-1467. https://doi.org/10.1061/(ASCE)0733-9445(1996)122:12(1459).
  29. Spiridonakos, M.D. and Chatzi, E.N. (2015), "Metamodeling of nonlinear structural systems with parametric uncertainty subject to stochastic dynamic excitation", Earthq. Struct., 8(4), 915-934. https://doi.org/10.12989/eas.2015.8.4.91.
  30. Towashiraporn, P., Duenas-Osorio, L., Craig, J.I. and Goodno, B.J. (2008), "An application of the response surface metamodel in building seismic fragility estimation", Proceedings of the 14th World Conference on Earthquake Engineering, Beijing, China, October.
  31. Ujjwalkarn, U. (2016), A Quick Introduction to Neural Networks; The Data Science Blog. https://ujjwalkarn.me/2016/08/09/quick-intro-neural-networks/
  32. Yang, T., Yuan, X., Zhong, J. and Yuan, W. (2020), "Near-fault pulse seismic ductility spectra for bridge columns based on machine learning", Soil Dyn. Earthq. Eng., 164, 107582. https://doi.org/10.1016/j.soildyn.2022.107582.
  33. Wong, J. (2020), Group Normalization - Jason Wong, Medium, San Francisco, CA, USA. https://jwong853.medium.com/group-normalization-explained-d52edbc6f062
  34. Wu, Y. and He, K. (2020), "Group normalization", Int. J. Comput. Vision, 128(3), 742-755. https://doi.org/10.1007/s11263-019-01198-w.
  35. Xie, Y., Zhang, J., DesRoches, R. and Padgett, J.E. (2019), "Seismic fragilities of single-column highway bridges with rocking column-footing", Earthq. Eng. Struct. Dyn., 48(7), 843-864. https://doi.org/10.1002/eqe.3164.
  36. Zhong, J., Zhu, Y., Zheng, X. and Han, Q. (2023), "Multivariable probabilistic seismic demand models for parametric fragility prediction of isolated bridges portfolios under pulse-like GMs", Eng. Struct., 292, 116517. https://doi.org/10.1016/j.engstruct.2023.116517.