Data‐driven modeling and prediction on hysteresis behavior of flexure RC columns using deep learning networks

Hysteresis behavior of structural components has been one of the research focus for the structural engineering community for decades, comprehensively determines the structural performance and safety, and plays an important role in structural disaster mitigation. It is of great significance to continuously monitor structural responses and accurately characterize structural hysteresis. Currently, the nonlinear properties of real‐world structural components cannot be obtained before its failure. Thus, a historical database is collected firstly. Then, a data‐driven analysis method is proposed for predicting hysteresis behaviors of reinforced concrete (RC) columns. A bidirectional LSTM (BLSTM) network is developed to model and predict hysteresis curves. The data with unfixed lengths are specially processed to simultaneously guarantee a uniform size and avoid data loss, and the clipping layers are inserted in the model to clip off inferior predictions and improve the accuracy. This methodology is systematically studied and validated by employing a sythetic database generated by numerical simulation and the full‐scale experiment database named PEER database. Result shows that proposed BLSTM can predict the hysteresis curves of the RC components with acceptable accuracy and robustness. Moreover, the interpretability analysis is performed on identifying the learning and prediction principle of the BLSTM model on hysteresis prediction and its accuracy variation under different model architectures.

[1]  Jixing Cao,et al.  Monitor wind characteristics and wind‐induced responses of a skyscraper during two typhoons , 2022, The Structural Design of Tall and Special Buildings.

[2]  Hanqing Zhang,et al.  Damage tracking and evaluation of RC columns with structural performances by using seismic monitoring data , 2022, Bulletin of Earthquake Engineering.

[3]  Jiazeng Shan,et al.  Feasibility of using self‐sensing component and response prediction model for rotation monitoring of shear wall structures , 2021, The Structural Design of Tall and Special Buildings.

[4]  J. Jeon,et al.  Seismic damage state predictions of reinforced concrete structures using stacked long short-term memory neural networks , 2021, Journal of Building Engineering.

[5]  Yongchao Yang,et al.  Hierarchical deep learning for data-driven identification of reduced-order models of nonlinear dynamical systems , 2021, Nonlinear Dynamics.

[6]  Jiazeng Shan,et al.  Residual‐based damage evaluation of RC columns using feature classification and model optimization , 2021, The Structural Design of Tall and Special Buildings.

[7]  Shanwu Li,et al.  A recurrent neural network framework with an adaptive training strategy for long-time predictive modeling of nonlinear dynamical systems , 2021 .

[8]  M. Limongelli,et al.  The value of structural health monitoring in seismic emergency management of bridges , 2020, Structure and Infrastructure Engineering.

[9]  Yongchao Yang,et al.  CNN-LSTM deep learning architecture for computer vision-based modal frequency detection , 2020 .

[10]  Weifeng Liu,et al.  Detecting structural damage under unknown seismic excitation by deep convolutional neural network with wavelet-based transmissibility data , 2020 .

[11]  Ruiyang Zhang,et al.  Deep long short-term memory networks for nonlinear structural seismic response prediction , 2019, Computers & Structures.

[12]  Hyo Seon Park,et al.  Convolutional neural network‐based wind‐induced response estimation model for tall buildings , 2019, Comput. Aided Civ. Infrastructure Eng..

[13]  Yong Huang,et al.  Multitask Sparse Bayesian Learning with Applications in Structural Health Monitoring , 2018, Comput. Aided Civ. Infrastructure Eng..

[14]  Faramarz Khoshnoudian,et al.  Experimental validation of a deep neural network—Sparse representation classification ensemble method , 2018, The Structural Design of Tall and Special Buildings.

[15]  B. Liu,et al.  Cumulative seismic damage assessment of reinforced concrete columns through cyclic and pseudo‐dynamic tests , 2017 .

[16]  Bin Xu,et al.  Data‐Based Model‐Free Hysteretic Restoring Force and Mass Identification for Dynamic Systems , 2015, Comput. Aided Civ. Infrastructure Eng..

[17]  Manolis Papadrakakis,et al.  Neural network based prediction schemes of the non-linear seismic response of 3D buildings , 2012, Adv. Eng. Softw..

[18]  Amr S. Elnashai,et al.  A new neural network‐based model for hysteretic behavior of materials , 2008 .

[19]  Dionisio Bernal,et al.  A data‐driven methodology for assessing impact of earthquakes on the health of building structural systems , 2006 .

[20]  S. Hochreiter,et al.  Long Short-Term Memory , 1997, Neural Computation.

[21]  Ming Yang,et al.  Bidirectional Long Short-Term Memory Networks for Relation Classification , 2015, PACLIC.

[22]  Ole Winther,et al.  Artificial Neural Networks for Nonlinear Dynamic Response Simulation in Mechanical Systems , 2011 .