pith. sign in

arxiv: 2411.02776 · v1 · pith:72EPT245new · submitted 2024-11-05 · 💻 cs.LG · stat.AP

Deep learning-based modularized loading protocol for parameter estimation of Bouc-Wen class models

Sebin Oh , Junho Song , Taeyong Kim This is my paper

Pith reviewed 2026-05-23 18:07 UTC · model grok-4.3

classification 💻 cs.LG stat.AP
keywords Bouc-Wen modelparameter estimationCNNhysteresis modelingloading protocolmodular deep learningstructural analysisdeep learning
0
0 comments X

The pith

A modular protocol using three separate CNNs constructs minimal loading histories to estimate Bouc-Wen model parameters faster without losing accuracy.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper introduces a modularized deep learning protocol for estimating parameters in Bouc-Wen class hysteresis models. It splits the problem into three independent modules covering basic hysteresis, degradation, and pinching, each with its own CNN trained on tailored minimal loading sequences. These modules combine into an optimal loading history, and the CNNs act as quick estimators. Tests on multi-story frames show reduced analysis time with maintained or better accuracy. The approach is designed to extend to other hysteresis models.

Core claim

The protocol decomposes optimal loading history construction and CNN-based parameter estimation into three independent sub-modules for basic hysteresis, structural degradation, and pinching effect. Three CNN architectures are trained on diverse loading histories to identify minimal loading history modules, which are combined for the full optimal history. The trained CNNs then serve as rapid estimators, and numerical evaluations on a 3-story steel frame and RC frame confirm significant reduction in total analysis time with maintained or improved accuracy.

What carries the argument

Three independent CNN architectures trained on loading history modules for basic hysteresis, degradation, and pinching, combined to recover full Bouc-Wen parameters.

If this is right

  • Reduces total analysis time in nonlinear time history analysis of a 3-story steel moment frame.
  • Maintains or improves estimation accuracy during fragility curve construction for a 3-story reinforced concrete frame.
  • Adapts to diverse hysteresis models through the modular sub-module design.
  • Offers a systematic approach for identifying general hysteresis models beyond the Bouc-Wen class.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The modular CNN setup could support incremental updates by retraining only one module when new data for a specific behavior arrives.
  • This decomposition might help apply the method to real-time structural monitoring where loading data arrives in segments.
  • Similar modularization could be tested on other nonlinear models with path-dependent behaviors in mechanics or materials science.

Load-bearing premise

The three independent CNN architectures, each trained only on its own loading history module, can be combined without loss of accuracy to recover the full set of Bouc-Wen parameters for any combination of basic hysteresis, degradation, and pinching.

What would settle it

Running the modular three-CNN system and a single integrated CNN on identical validation cases of Bouc-Wen responses and checking whether the modular estimates have comparable or lower error rates.

Figures

Figures reproduced from arXiv: 2411.02776 by Junho Song, Sebin Oh, Taeyong Kim.

Figure 1
Figure 1. Figure 1: Overall framework to develop a modularized loading protocol. BSC, DGD, and PCH denote three hysteresis categories: basic hysteresis, structural degradation, and the pinching effect. 2. Review: modified Bouc-Wen-Baber-Noori model Kim et al. [28] proposed the modified Bouc-Wen-Baber-Noori (m-BWBN) model to enhance the practicality of the Bouc-Wen-Baber-Noori (BWBN) model [29, 30], by rearranging the mathemat… view at source ↗
Figure 2
Figure 2. Figure 2: Variations of the hysteresis curves for three different values of the BSC parameters [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Variations of the hysteresis curves for three different values of the DGD parameters [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Variations of the hysteresis curves for three different values of the PCH parameters. 6 [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Loading history used for the CNN architecture development. 3.2.1. Parameteres for basic hysteresis and degradation [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: CNN architecture for the BSC and DGD parameters. (a) (b) [PITH_FULL_IMAGE:figures/full_fig_p009_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Performance of the trained models demonstrated by correlation coefficients of (a) BSC parameters and (b) DGD parameters, respectively. 3.2.2. Pinching parameters The pinching effect modulates the stiffness, particularly in regions near zero displacement, and the PCH parameters primarily govern the spatial correlation between small- and large-displacement domains. The CNN architecture is designed to capture… view at source ↗
Figure 8
Figure 8. Figure 8: The CNN architecture for the PCH parameters. The input length d depends on which loading history is applied. For the PCH parameters, n1 = 1024, n2 = 256, n3 = 64, n4 = 16 are used, and the number of parameters nparam = 6. Similar to the BSC and DGD parameters, the CNN architecture of [PITH_FULL_IMAGE:figures/full_fig_p010_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Performance of the trained models demonstrated by correlation coefficients of the PCH parameters. As shown in [PITH_FULL_IMAGE:figures/full_fig_p010_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Collective performance of trained CNN models: (a) the reference loading history and (b) the new loading history. To further investigate the performance of the CNN models, we present a scatter plot in [PITH_FULL_IMAGE:figures/full_fig_p011_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Scatter plots of the true and predicted hysteresis energy for 50,000 parameter sets of the test data under two different loading histories: (a) the reference loading history and (b) the new loading history. 4. Loading history modules for different hysteresis categories The CNN architectures presented in Section 3 are now trained on different loading histories to identify the loading history module for eac… view at source ↗
Figure 12
Figure 12. Figure 12: Loading history analysis for the BSC parameters: correlation coefficients between the true and predicted values for 50,000 test data using the CNN models trained with different loading histories [PITH_FULL_IMAGE:figures/full_fig_p014_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Loading history analysis for the DGD parameters: correlation coefficients between the true and predicted values for 50,000 test data using the CNN models trained with different loading histories. 4.4. Pinching parameters [PITH_FULL_IMAGE:figures/full_fig_p014_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Loading history analysis for the PCH parameters: correlation coefficients between the true and predicted values for 50,000 test data using the CNN models trained with different loading histories. 4.5. Proposed loading protocol [PITH_FULL_IMAGE:figures/full_fig_p015_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Proposed modularized loading protocol. Furthermore, to facilitate the practical use of the protocol, we propose optimal loading histories for four different BW class models: (1) BW, (2) BW with degradation, (3) BWBN, and (4) m-BWBN models. Note that the BW model consists of only the BSC parameters, while the BW model with degradation includes the BSC and DGD parameters. The BWBN and m-BWBN models involve … view at source ↗
Figure 16
Figure 16. Figure 16: Examples of optimal loading histories for different Bouc-Wen class models. 5. Numerical investigations Three numerical investigations are performed to evaluate the efficiency and effectiveness of the proposed protocol. First, the effectiveness of the optimal loading history is demonstrated by comparing the parameter estimation accuracy obtained from the optimal and reference loading histories using a gene… view at source ↗
Figure 17
Figure 17. Figure 17: illustrates the pushover curve for the 3-story SAC structure, where the base shear force is normalized by the mass of the structure. The yield displacement uy is estimated to be 14 cm following the method outlined in Section 4.5 [PITH_FULL_IMAGE:figures/full_fig_p018_17.png] view at source ↗
Figure 18
Figure 18. Figure 18: Comparison between the hysteresis curves for the (a) BW and (b) m-BWBN models, obtained using the OpenSees (True), the genetic algorithm (GA), and the CNN models (CNN). The corresponding optimal loading histories are used, as shown in [PITH_FULL_IMAGE:figures/full_fig_p019_18.png] view at source ↗
Figure 19
Figure 19. Figure 19: A pushover curve for the 3-story RC structure. The yield displacement uy is obtained as 4.4 cm [PITH_FULL_IMAGE:figures/full_fig_p020_19.png] view at source ↗
Figure 20
Figure 20. Figure 20: compares the hysteresis curve from OpenSees with the curves obtained using the parameters estimated by the two different methods. The fitness of the predicted hysteresis curves to the OpenSees curve is limited compared to the case of the 3-story SAC structure in [PITH_FULL_IMAGE:figures/full_fig_p020_20.png] view at source ↗
Figure 21
Figure 21. Figure 21: Fragility curves for three different damage states (DSs) obtained using the OpenSees and the m￾BWBN model with the parameters estimated from the CNN models (CNN) and the genetic algorithm (GA) [PITH_FULL_IMAGE:figures/full_fig_p021_21.png] view at source ↗
read the original abstract

This study proposes a modularized deep learning-based loading protocol for optimal parameter estimation of Bouc-Wen (BW) class models. The protocol consists of two key components: optimal loading history construction and CNN-based rapid parameter estimation. Each component is decomposed into independent sub-modules tailored to distinct hysteretic behaviors-basic hysteresis, structural degradation, and pinching effect-making the protocol adaptable to diverse hysteresis models. Three independent CNN architectures are developed to capture the path-dependent nature of these hysteretic behaviors. By training these CNN architectures on diverse loading histories, minimal loading sequences, termed \textit{loading history modules}, are identified and then combined to construct an optimal loading history. The three CNN models, trained on the respective loading history modules, serve as rapid parameter estimators. Numerical evaluation of the protocol, including nonlinear time history analysis of a 3-story steel moment frame and fragility curve construction for a 3-story reinforced concrete frame, demonstrates that the proposed protocol significantly reduces total analysis time while maintaining or improving estimation accuracy. The proposed protocol can be extended to other hysteresis models, suggesting a systematic approach for identifying general hysteresis models.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript proposes a modularized deep learning-based loading protocol for parameter estimation of Bouc-Wen class models. The protocol decomposes loading histories into three independent modules (basic hysteresis, structural degradation, pinching) and trains separate CNN architectures on each to identify minimal loading sequences and serve as rapid estimators; these are then combined for full parameter recovery. Numerical evaluations via nonlinear time-history analysis of a 3-story steel moment frame and fragility curves for a 3-story RC frame are cited to claim significant reduction in total analysis time while maintaining or improving accuracy, with the method positioned as extensible to other hysteresis models.

Significance. If the modular CNN combination recovers accurate parameters even under simultaneous variation of all three behavior classes, the protocol would provide a systematic, time-efficient alternative to traditional identification methods for path-dependent hysteretic models, with clear utility in structural dynamics applications.

major comments (2)
  1. [Abstract (protocol description and numerical evaluation paragraph)] The central claim that the three independently trained CNNs can be concatenated to recover the full Bouc-Wen parameter vector for arbitrary joint combinations of basic hysteresis, degradation, and pinching rests on an unverified separability assumption. Bouc-Wen-class models are defined by coupled differential equations in which degradation and pinching coefficients directly modulate the restoring-force evolution; training distributions that vary only one behavior class at a time therefore leave the combined estimator untested on the exact regime the protocol is intended to handle.
  2. [Abstract] No quantitative metrics, error bars, baseline comparisons, training/test split details, or description of how joint-variation test cases were generated appear in support of the accuracy claim. The abstract states only that the protocol 'maintains or improves estimation accuracy,' rendering the numerical-evaluation paragraph unverifiable from the provided text.
minor comments (1)
  1. [Abstract] The abstract refers to 'three independent CNN architectures' and 'minimal loading sequences, termed loading history modules' without defining the precise input representation (e.g., time-series length, normalization) or output parameterization used by each network.

Simulated Author's Rebuttal

2 responses · 0 unresolved

Thank you for the referee's insightful comments. We provide point-by-point responses to the major comments and indicate the revisions we will make to address them.

read point-by-point responses

  1. Referee: [Abstract (protocol description and numerical evaluation paragraph)] The central claim that the three independently trained CNNs can be concatenated to recover the full Bouc-Wen parameter vector for arbitrary joint combinations of basic hysteresis, degradation, and pinching rests on an unverified separability assumption. Bouc-Wen-class models are defined by coupled differential equations in which degradation and pinching coefficients directly modulate the restoring-force evolution; training distributions that vary only one behavior class at a time therefore leave the combined estimator untested on the exact regime the protocol is intended to handle.

    Authors: We agree that explicit verification on joint variations is important to fully substantiate the separability assumption. While the numerical evaluations in Sections 4 and 5 involve complete Bouc-Wen models with all parameters varying simultaneously in the context of the structural analyses, the training process for the CNNs was indeed modular. In the revised manuscript, we will include additional experiments where all three behavior classes vary jointly in the test cases to directly validate the combined estimator. revision: yes

  2. Referee: [Abstract] No quantitative metrics, error bars, baseline comparisons, training/test split details, or description of how joint-variation test cases were generated appear in support of the accuracy claim. The abstract states only that the protocol 'maintains or improves estimation accuracy,' rendering the numerical-evaluation paragraph unverifiable from the provided text.

    Authors: Due to space limitations in the abstract, detailed metrics were not included. We will revise the abstract to incorporate key quantitative results, such as mean absolute percentage errors for parameter estimation, comparisons with traditional methods, and a brief note on the test set composition, while maintaining the abstract's conciseness. revision: yes

Circularity Check

0 steps flagged

No significant circularity; standard CNN application with empirical validation

full rationale

The protocol decomposes loading histories into modules and trains independent CNNs for each hysteretic behavior class, then concatenates the estimators. Accuracy and time-reduction claims rest on numerical evaluations (nonlinear time-history analysis of steel frames and fragility curves for RC frames), which are external to the training procedure and not forced by construction. No equations equate a reported prediction to a fitted input, no self-citation chains justify core premises, and no ansatz or uniqueness result is smuggled in. The method is a direct application of supervised learning to a new domain; the reader's noted assumption about module independence is a correctness concern, not a circularity reduction.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

The abstract provides no explicit free parameters, axioms, or invented entities beyond standard assumptions of supervised learning and the existence of Bouc-Wen class models. All technical details required to audit the ledger are absent from the supplied text.

pith-pipeline@v0.9.0 · 5726 in / 1153 out tokens · 28157 ms · 2026-05-23T18:07:18.348709+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

56 extracted references · 56 canonical work pages

  1. [1]

    The hysteresis Bouc-Wen model, a survey.Archives of Compu- tational Methods in Engineering , 16(2):161–188, 1 2009

    Mohammed Ismail, Fay¸ cal Ikhouane, and Jos´ e Rodellar. The hysteresis Bouc-Wen model, a survey.Archives of Compu- tational Methods in Engineering , 16(2):161–188, 1 2009. ISSN 11343060. doi: 10.1007/s11831-009-9031-8

    work page doi:10.1007/s11831-009-9031-8 2009

  2. [2]

    Generalized Bouc–Wen Model for Highly Asymmetric Hysteresis

    Junho Song and Armen Der Kiureghian. Generalized Bouc–Wen Model for Highly Asymmetric Hysteresis. Journal of Engineering Mechanics, 132(6):610–618, 6 2006. ISSN 0733-9399. doi: 10.1061/(asce)0733-9399(2006)132:6(610)

    work page doi:10.1061/(asce)0733-9399(2006)132:6(610 2006

  3. [3]

    Nondimensionalized Bouc–Wen model with structural degradation for Kalman fil- ter–based real-time monitoring

    Sung Yong Kim and Cheol Ho Lee. Nondimensionalized Bouc–Wen model with structural degradation for Kalman fil- ter–based real-time monitoring. Engineering Structures , 244, 10 2021. ISSN 18737323. doi: 10.1016/j.engstruct.2021. 112674

    work page doi:10.1016/j.engstruct.2021 2021

  4. [4]

    R. Bouc. Forced vibrations of mechanical systems with hysteresis. Proc. of the Fourth Conference on Nonlinear Oscilla- tions, Prague, 1967 , 1967. URL https://cir.nii.ac.jp/crid/1570291225556669696.bib?lang=en

    work page arXiv 1967

  5. [5]

    F. Ma, H. Zhang, A. Bockstedte, G. C. Foliente, and P. Paevere. Parameter analysis of the differential model of hysteresis. Journal of Applied Mechanics, Transactions ASME , 71(3):342–349, 5 2004. ISSN 00218936. doi: 10.1115/1.1668082

    work page doi:10.1115/1.1668082 2004

  6. [6]

    Response prediction of nonlinear hysteretic systems by deep neural networks

    Taeyong Kim, Oh Sung Kwon, and Junho Song. Response prediction of nonlinear hysteretic systems by deep neural networks. Neural Networks, 111:1–10, 3 2019. ISSN 18792782. doi: 10.1016/j.neunet.2018.12.005

    work page doi:10.1016/j.neunet.2018.12.005 2019

  7. [7]

    Nonlinear estimation of the Bouc-Wen model with parameter boundaries: Application to seismic isolators

    Vincenzo Niola, Gianluca Palli, Salvatore Strano, and Mario Terzo. Nonlinear estimation of the Bouc-Wen model with parameter boundaries: Application to seismic isolators. Computers and Structures , 222:1–9, 10 2019. ISSN 00457949. doi: 10.1016/j.compstruc.2019.06.006

    work page doi:10.1016/j.compstruc.2019.06.006 2019

  8. [8]

    A. E. Charalampakis and V. K. Koumousis. Identification of Bouc-Wen hysteretic systems by a hybrid evolutionary algorithm. Journal of Sound and Vibration , 314(3-5):571–585, 7 2008. ISSN 0022460X. doi: 10.1016/j.jsv.2008.01.018

    work page doi:10.1016/j.jsv.2008.01.018 2008

  9. [9]

    Seismic behaviors of precast assembled bridge columns connected with prestressed threaded steel bar: Experimental test and hysteretic model

    Qifang Xie, Xudong Zhao, Xiaofei Yao, Wenming Hao, and Fangzheng Hu. Seismic behaviors of precast assembled bridge columns connected with prestressed threaded steel bar: Experimental test and hysteretic model. Advances in Structural Engineering, 23(9):1975–1988, 7 2020. ISSN 20484011. doi: 10.1177/1369433220903988

    work page doi:10.1177/1369433220903988 1975

  10. [10]

    Hysteretic model and parameter identification of RC bridge piers based on a new modified Bouc-Wen model

    Kelun Wei and Yazhou Xu. Hysteretic model and parameter identification of RC bridge piers based on a new modified Bouc-Wen model. Structures, 43:1766–1777, 9 2022. ISSN 23520124. doi: 10.1016/j.istruc.2022.07.049

    work page doi:10.1016/j.istruc.2022.07.049 2022

  11. [11]

    Guide for testing reinforced concrete structural elements under slowly applied simulated seismic loads

    ACI Committee 374. Guide for testing reinforced concrete structural elements under slowly applied simulated seismic loads. Technical report, American Concrete Institute, Farmington Hills, MI, 8 2013. 22

    work page 2013

  12. [12]

    Guidelines for Cyclic Seismic Testing of Components of Steel Structures

    ATC-24. Guidelines for Cyclic Seismic Testing of Components of Steel Structures. Technical report, Applied Technology Council, Redwood City, CA, 1992

    work page 1992

  13. [13]

    Interim Testing Protocols for Determining the Seismic Performance Characteristics of Structural and Nonstructural Components

    FEMA. Interim Testing Protocols for Determining the Seismic Performance Characteristics of Structural and Nonstructural Components. Technical report, FEMA 461, Washington, DC, 2007. URL www.ATCouncil.org

    work page 2007

  14. [14]

    Development of a Testing Protocol for Woodframe Structures

    Helmut Krawinkler, Francisco Parisi, Luis Ibarra, Ashraf Ayoub, and Ricardo Medina. Development of a Testing Protocol for Woodframe Structures. Technical report, CUREE, Richmond, CA, 2001. URL https://www.researchgate.net/ publication/245911208

    work page arXiv 2001

  15. [15]

    Bouc–Wen class models considering hysteresis mechanism of RC columns in nonlinear dynamic analysis

    Sebin Oh, Taeyong Kim, and Junho Song. Bouc–Wen class models considering hysteresis mechanism of RC columns in nonlinear dynamic analysis. International Journal of Non-Linear Mechanics , 148:104263, 1 2023. ISSN 0020-7462. doi: 10.1016/J.IJNONLINMEC.2022.104263

    work page doi:10.1016/j.ijnonlinmec.2022.104263 2023

  16. [16]

    Experimental study on hysteretic behavior of structural stainless steels under cyclic loading

    Feng Zhou and Lu Li. Experimental study on hysteretic behavior of structural stainless steels under cyclic loading. Journal of Constructional Steel Research , 122:94–109, 7 2016. ISSN 0143974X. doi: 10.1016/j.jcsr.2016.03.006

    work page doi:10.1016/j.jcsr.2016.03.006 2016

  17. [17]

    Variations in the hysteretic behavior of LRBs as a function of applied loading

    G¨ okhan¨Ozdemir, Beyhan Bayhan, and Polat G¨ ulkan. Variations in the hysteretic behavior of LRBs as a function of applied loading. Structural Engineering and Mechanics , 67(1):69–78, 7 2018. ISSN 15986217. doi: 10.12989/sem.2018.67.1.069

    work page doi:10.12989/sem.2018.67.1.069 2018

  18. [18]

    Effects of Loading Protocol on the Cyclic Response of Woodframe Shearwalls

    Kip Gatto and Chia-Ming Uang. Effects of Loading Protocol on the Cyclic Response of Woodframe Shearwalls. Journal of Structural Engineering , 129(10):1384–1393, 2003. doi: 10.1061/ASCE0733-94452003129:101384

    work page doi:10.1061/asce0733-94452003129:101384 2003

  19. [19]

    Chang Seok Lee and Sang Whan Han. An Accurate Numerical Model Simulating Hysteretic Behavior of Reinforced Concrete Columns Irrespective of Types of Loading Protocols.International Journal of Concrete Structures and Materials , 15(1), 12 2021. ISSN 22341315. doi: 10.1186/s40069-020-00446-5

    work page doi:10.1186/s40069-020-00446-5 2021

  20. [20]

    Foschi, and Frank Lam

    Minghao Li, Ricardo O. Foschi, and Frank Lam. Modeling Hysteretic Behavior of Wood Shear Walls with a Protocol- Independent Nail Connection Algorithm. Journal of Structural Engineering , 138(1):99–108, 1 2012. ISSN 0733-9445. doi: 10.1061/(asce)st.1943-541x.0000438

    work page doi:10.1061/(asce)st.1943-541x.0000438 2012

  21. [21]

    Adam Zsarn´ oczay and Jack W. Baker. Using model error in response history analysis to evaluate component calibration methods. Earthquake Engineering and Structural Dynamics , 49(2):175–193, 2 2020. ISSN 10969845. doi: 10.1002/eqe.3234

    work page doi:10.1002/eqe.3234 2020

  22. [22]

    Masset, R

    Gilberto A. Ortiz, Diego A. Alvarez, and Daniel Bedoya-Ru´ ız. Identification of Bouc-Wen type models using multi- objective optimization algorithms. Computers and Structures , 114-115:121–132, 1 2013. ISSN 00457949. doi: 10.1016/j. compstruc.2012.10.016

    work page doi:10.1016/j 2013

  23. [23]

    O’Reilly Media, second edition, 2017

    Aure´ elien Ge´ eron.Hands-on Machine Learning with Scikit-Learn, Keras, and TensorFlow: Concepts, Tools, and Tech- niques to Build Intelligent Systems . O’Reilly Media, second edition, 2017

    work page 2017

  24. [24]

    A Guide to Convolutional Neural Networks for Computer Vision

    Salman Khan, Hossein Rahmani, Syed Afaq Ali Shah, and Mohammed Bennamoun. A Guide to Convolutional Neural Networks for Computer Vision . Springer Cham, 1 edition, 2018. doi: https://doi.org/10.1007/978-3-031-01821-3

    work page doi:10.1007/978-3-031-01821-3 2018

  25. [25]

    Humaidi, Ayad Al-Dujaili, Ye Duan, Omran Al-Shamma, J

    Laith Alzubaidi, Jinglan Zhang, Amjad J. Humaidi, Ayad Al-Dujaili, Ye Duan, Omran Al-Shamma, J. Santamar´ ıa, Mohammed A. Fadhel, Muthana Al-Amidie, and Laith Farhan. Review of deep learning: concepts, CNN archi- tectures, challenges, applications, future directions. Journal of Big Data , 8(1), 12 2021. ISSN 21961115. doi: 10.1186/s40537-021-00444-8

    work page doi:10.1186/s40537-021-00444-8 2021

  26. [26]

    Deep Residual Learning for Image Recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep Residual Learning for Image Recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition , pages 770–778, 2016. URL http://image-net.org/ challenges/LSVRC/2015/

    work page 2016

  27. [27]

    Imagenet: A large-scale hierarchical image database

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. Imagenet: A large-scale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition , pages 248–255. IEEE, 6 2009. ISBN 9781424439911

    work page 2009

  28. [28]

    Deep learning based seismic response prediction of hysteretic systems having degradation and pinching

    Taeyong Kim, Oh Sung Kwon, and Junho Song. Deep learning based seismic response prediction of hysteretic systems having degradation and pinching. Earthquake Engineering and Structural Dynamics , 52(8):2384–2406, 7 2023. ISSN 10969845. doi: 10.1002/eqe.3796

    work page doi:10.1002/eqe.3796 2023

  29. [29]

    RANDOM VIBRATION OF DEGRADING, PINCHING SYSTEMS

    Thomas T Baber and Mohammad N Noori. RANDOM VIBRATION OF DEGRADING, PINCHING SYSTEMS. Journal of Engineering Mechanics , 111(8):977–1092, 8 1985. ISSN 0733-9399

    work page 1985

  30. [30]

    Baber and Mohammad N

    Thomas T. Baber and Mohammad N. Noori. Modeling general hysteresis behavior and random vibration application. Journal of Vibration and Acoustics , 108(4):411–420, 1986. ISSN 0733-9399. doi: 10.1061/(asce)0733-9399(1985)111: 8(1010)

    work page doi:10.1061/(asce)0733-9399(1985)111: 1986

  31. [31]

    Uncertainty quantification in the calibration of numerical elements in nonlinear seismic analysis

    Hongzhou Zhang, Oh Sung Kwon, and Constantin Christopoulos. Uncertainty quantification in the calibration of numerical elements in nonlinear seismic analysis. Earthquake Engineering & Structural Dynamics , 51(12):3000–3021, 10 2022. ISSN 1096-9845. doi: 10.1002/EQE.3711. URL https://onlinelibrary.wiley.com/doi/full/10.1002/eqe.3711https: //onlinelibrary.w...

    work page doi:10.1002/eqe.3711 2022

  32. [32]

    Hybrid-simulation-based model calibration method for nonlinear seismic analysis

    Hongzhou Zhang, Oh Sung Kwon, and Constantin Christopoulos. Hybrid-simulation-based model calibration method for nonlinear seismic analysis. Earthquake Engineering & Structural Dynamics , 53(3):1067–1084, 3 2024. ISSN 1096-9845. doi: 10.1002/EQE.4059. URL https://onlinelibrary.wiley.com/doi/full/10.1002/eqe.4059https://onlinelibrary. wiley.com/doi/abs/10....

    work page doi:10.1002/eqe.4059 2024

  33. [33]

    PEER Structural Performance Database User’s Manual (Version 1.2)

    Michael Berry, Myles Parrish, and Marc Eberhard. PEER Structural Performance Database User’s Manual (Version 1.2). Technical report, Pacific Earthquake Engineering Research Center, 1 2004

    work page 2004

  34. [34]

    Zele Li, Mohammad Noori, Ying Zhao, Chunfeng Wan, Decheng Feng, and Wael A. Altabey. A multi-objective optimization algorithm for Bouc–Wen–Baber–Noori model to identify reinforced concrete columns failing in different modes.Proceedings of the Institution of Mechanical Engineers, Part L: Journal of Materials: Design and Applications , 235(9):2165–2182, 10

    work page

  35. [35]

    doi: 10.1177/14644207211020028

    ISSN 20413076. doi: 10.1177/14644207211020028

    work page doi:10.1177/14644207211020028

  36. [36]

    Parameter Sensitivity Analysis and Identification of an Improved Symmetrical Hysteretic Model for RC Hollow Columns

    Huaping Yang, Jing Li, Changjiang Shao, Yongjiu Qian, Qiming Qi, and Jianxian He. Parameter Sensitivity Analysis and Identification of an Improved Symmetrical Hysteretic Model for RC Hollow Columns. Symmetry, 14(5), 5 2022. ISSN 20738994. doi: 10.3390/sym14050945. 23

    work page doi:10.3390/sym14050945 2022

  37. [37]

    Ajavakom, C

    N. Ajavakom, C. H. Ng, and F. Ma. Performance of nonlinear degrading structures: Identification, validation, and prediction. Computers and Structures , 86(7-8):652–662, 4 2008. ISSN 00457949. doi: 10.1016/j.compstruc.2007.07.014

    work page doi:10.1016/j.compstruc.2007.07.014 2008

  38. [38]

    P. C. Jennings, G. W. Housner, and N. C. Tsai. Simulated Earthquake Motions. Technical report, California Institute of Technology, Pasadena, California, 4 1968

    work page 1968

  39. [39]

    PRESTANDARD AND COMMENTARY FOR THE SEISMIC REHABILITATION OF BUILDINGS

    FEMA 356. PRESTANDARD AND COMMENTARY FOR THE SEISMIC REHABILITATION OF BUILDINGS. Tech- nical report, Federal Emergency Management Agency (FEMA), Washington, DC, 11 2000

    work page 2000

  40. [40]

    De Luca, D

    F. De Luca, D. Vamvatsikos, and I. Iervolino. Near-optimal piecewise linear fits of static pushover capacity curves for equivalent SDOF analysis. Earthquake Engineering and Structural Dynamics , 42(4):523–543, 4 2013. ISSN 10969845. doi: 10.1002/eqe.2225

    work page doi:10.1002/eqe.2225 2013

  41. [41]

    Hysteresis modeling of reinforced concrete structures: State of the art

    Piyali Sengupta and Bing Li. Hysteresis modeling of reinforced concrete structures: State of the art. ACI Structural Journal, 114(1):25–38, 1 2017. ISSN 08893241. doi: 10.14359/51689422

    work page doi:10.14359/51689422 2017

  42. [42]

    Regional seismic fragility of bridge network derived by covariance matrix model of bridge portfolios

    Jian Zhong, Sien Zhou, Hao Wang, and Huimin Hu. Regional seismic fragility of bridge network derived by covariance matrix model of bridge portfolios. Engineering Structures, 309, 6 2024. ISSN 18737323. doi: 10.1016/j.engstruct.2024. 118035

    work page doi:10.1016/j.engstruct.2024 2024

  43. [43]

    Regional-scale nonlinear structural seismic response prediction by neural network

    Zekun Xu, Jun Chen, Jiaxu Shen, and Mengjie Xiang. Regional-scale nonlinear structural seismic response prediction by neural network. Engineering Failure Analysis , 154, 12 2023. ISSN 13506307. doi: 10.1016/j.engfailanal.2023.107707

    work page doi:10.1016/j.engfailanal.2023.107707 2023

  44. [44]

    Steelman and Jerome F

    Joshua S. Steelman and Jerome F. Hajjar. Influence of inelastic seismic response modeling on regional loss estimation. Engineering Structures, 31(12):2976–2987, 12 2009. ISSN 01410296. doi: 10.1016/j.engstruct.2009.07.026

    work page doi:10.1016/j.engstruct.2009.07.026 2009

  45. [45]

    NHERI-SimCenter/R2DTool: Version 4.0.0, 1 2024

    Frank McKenna, Stevan Gavrilovic, Adam Zsarnoczay, Jinyan Zhao, Kuanshi Zhong, Barbaros Cetiner, Sang-ri Yi, Wael Elhaddad, and Pedro Arduino. NHERI-SimCenter/R2DTool: Version 4.0.0, 1 2024. URL https://doi.org/10.5281/ zenodo.10448043

    work page 2024

  46. [46]

    SEISMIC DEMANDS FOR PERFORMANCE EVALUATION OF STEEL MO- MENT RESISTING FRAME STRUCTURES

    Akshay Gupta and Helmut Krawinkler. SEISMIC DEMANDS FOR PERFORMANCE EVALUATION OF STEEL MO- MENT RESISTING FRAME STRUCTURES. Technical report, The John A. Blume Earthquake Engineering Center,

    work page

  47. [47]

    URL http://blume.stanford.edu

    work page

  48. [48]

    Goel and Anil K

    Rakesh K. Goel and Anil K. Chopra. Evaluation of Modal and FEMA Pushover Analyses: SAC Buildings, 2004. ISSN 87552930

    work page 2004

  49. [49]

    SUMMARY OF SAC CASE STUDY BUILDING ANALYSES.Journal of performance of constructed facilities, 12(4):202–212, 1998

    Gregory G Deierlein. SUMMARY OF SAC CASE STUDY BUILDING ANALYSES.Journal of performance of constructed facilities, 12(4):202–212, 1998

    work page 1998

  50. [50]

    Energy-based fragility curves of building structures equipped with viscous dampers

    Ying Zhou, Yi Xiao, and Mohammed Samier Sebaq. Energy-based fragility curves of building structures equipped with viscous dampers. Structures, 44:1660–1679, 10 2022. ISSN 23520124. doi: 10.1016/j.istruc.2022.08.101

    work page doi:10.1016/j.istruc.2022.08.101 2022

  51. [51]

    OpenSees: A Framework for Earthquake Engineering Simulation

    Frank McKenna. OpenSees: A Framework for Earthquake Engineering Simulation. Computing in Science & Engineering , 13(4):58–66, 2011

    work page 2011

  52. [52]

    The effect of material and ground motion uncertainty on the seismic vulnerability curves of RC structure

    Oh Sung Kwon and Amr Elnashai. The effect of material and ground motion uncertainty on the seismic vulnerability curves of RC structure. Engineering Structures, 28(2):289–303, 1 2006. ISSN 01410296. doi: 10.1016/j.engstruct.2005.07.010

    work page doi:10.1016/j.engstruct.2005.07.010 2006

  53. [53]

    NGA project strong-motion database

    Brian Chiou, Robert Darragh, Nick Gregor, and Walter Silva. NGA project strong-motion database. Earthquake Spectra, 24(1):23–44, 2008. ISSN 87552930. doi: 10.1193/1.2894831

    work page doi:10.1193/1.2894831 2008

  54. [54]

    An overview of the NGA project

    Maurice Power, Brian Chiou, Norman Abrahamson, Yousef Bozorgnia, Thomas Shantz, and Clifford Roblee. An overview of the NGA project. Earthquake Spectra, 24(1):3–21, 2008. ISSN 87552930. doi: 10.1193/1.2894833

    work page doi:10.1193/1.2894833 2008

  55. [55]

    Probabilistic evaluation of seismic responses using deep learning method

    Taeyong Kim, Junho Song, and Oh Sung Kwon. Probabilistic evaluation of seismic responses using deep learning method. Structural Safety, 84, 5 2020. ISSN 01674730. doi: 10.1016/j.strusafe.2019.101913

    work page doi:10.1016/j.strusafe.2019.101913 2020

  56. [56]

    On Information and Sufficiency

    S Kullback and R A Leibler. On Information and Sufficiency. The annals of mathematical statistics , 22(1):79–86, 1951. Appendix A. Training and test errors for CNN models Table A.6 summarizes the training and test errors for the CNN models trained on the reference loading history (denoted as Ref. in the table), providing a detailed demonstration of the ef...

    work page 1951