Physics-guided Convolutional Neural Network (PhyCNN) for Data-driven Seismic Response Modeling

Seismic events, among many other natural hazards, reduce due functionality and exacerbate vulnerability of in-service buildings. Accurate modeling and prediction of building's response subjected to earthquakes makes possible to evaluate building performance. To this end, we leverage the recent advances in deep learning and develop a physics-guided convolutional neural network (PhyCNN) framework for data-driven seismic response modeling and serviceability assessment of buildings. The proposed PhyCNN approach is capable of accurately predicting building's seismic response in a data-driven fashion without the need of a physics-based analytical/numerical model. The basic concept is to train a deep PhyCNN model based on available seismic input-output datasets (e.g., from simulation or sensing) and physics constraints. The trained PhyCNN can then used as a surrogate model for structural seismic response prediction. Available physics (e.g., the law of dynamics) can provide constraints to the network outputs, alleviate overfitting issues, reduce the need of big training datasets, and thus improve the robustness of the trained model for more reliable prediction. The trained surrogate model is then utilized for fragility analysis given certain limit state criteria (e.g., the serviceability state). In addition, an unsupervised learning algorithm based on K-means clustering is also proposed to partition the limited number of datasets to training, validation and prediction categories, so as to maximize the use of limited datasets. The performance of the proposed approach is demonstrated through three case studies including both numerical and experimental examples. Convincing results illustrate that the proposed PhyCNN paradigm outperforms conventional pure data-based neural networks.

[1]  James E. Braun,et al.  An Inverse Gray-Box Model for Transient Building Load Prediction , 2002 .

[2]  Brahim Chaib-draa,et al.  Parametric Exponential Linear Unit for Deep Convolutional Neural Networks , 2016, 2017 16th IEEE International Conference on Machine Learning and Applications (ICMLA).

[3]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[4]  Maziar Raissi,et al.  Deep Hidden Physics Models: Deep Learning of Nonlinear Partial Differential Equations , 2018, J. Mach. Learn. Res..

[5]  Stephen A. Billings,et al.  International Journal of Control , 2004 .

[6]  Stefan C. Kremer,et al.  Recurrent Neural Networks , 2013, Handbook on Neural Information Processing.

[7]  Lennart Ljung,et al.  Nonlinear black-box modeling in system identification: a unified overview , 1995, Autom..

[8]  K. Worden,et al.  Past, present and future of nonlinear system identification in structural dynamics , 2006 .

[9]  François M. Hemez,et al.  Uncertainty and Sensitivity Analysis of Damage Identification Results Obtained Using Finite Element Model Updating , 2009, Comput. Aided Civ. Infrastructure Eng..

[10]  Mohammad R. Jahanshahi,et al.  Evaluation of deep learning approaches based on convolutional neural networks for corrosion detection , 2018 .

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

[12]  Paris Perdikaris,et al.  Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations , 2019, J. Comput. Phys..

[13]  V. Ediger,et al.  ARIMA forecasting of primary energy demand by fuel in Turkey , 2007 .

[14]  Patrick Paultre,et al.  Damage Identification in a Truss Tower by Regularized Model Updating , 2010 .

[15]  Ruiyang Zhang,et al.  Sparse Bayesian learning for structural damage identification , 2020 .

[16]  Stephen Grossberg,et al.  Recurrent neural networks , 2013, Scholarpedia.

[17]  Li Zhou,et al.  An adaptive extended Kalman filter for structural damage identification , 2006 .

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

[19]  Ka-Veng Yuen,et al.  Efficient Bayesian sensor placement algorithm for structural identification: a general approach for multi‐type sensory systems , 2015 .

[20]  Geoffrey E. Hinton,et al.  Rectified Linear Units Improve Restricted Boltzmann Machines , 2010, ICML.

[21]  Paris Perdikaris,et al.  Physics-Constrained Deep Learning for High-dimensional Surrogate Modeling and Uncertainty Quantification without Labeled Data , 2019, J. Comput. Phys..

[22]  Zilong Wang,et al.  Automated damage-sensitive feature extraction using unsupervised convolutional neural networks , 2018, Smart Structures and Materials + Nondestructive Evaluation and Health Monitoring.

[23]  Ying Lei,et al.  Identification, Model Updating, and Response Prediction of an Instrumented 15-Story Steel-Frame Building , 2006 .

[24]  Ka-Veng Yuen Updating large models for mechanical systems using incomplete modal measurement , 2012 .

[25]  B. Irie,et al.  Capabilities of three-layered perceptrons , 1988, IEEE 1988 International Conference on Neural Networks.

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

[27]  Oral Büyüköztürk,et al.  Deep Learning‐Based Crack Damage Detection Using Convolutional Neural Networks , 2017, Comput. Aided Civ. Infrastructure Eng..

[28]  Padhraic Smyth,et al.  Clustering Using Monte Carlo Cross-Validation , 1996, KDD.

[29]  Jian Zhang,et al.  Novel support vector regression for structural system identification , 2007 .

[30]  Trupti M. Kodinariya,et al.  Review on determining number of Cluster in K-Means Clustering , 2013 .

[31]  Oral Büyüköztürk,et al.  Continuous Monitoring of High‐Rise Buildings Using Seismic InterferometryContinuous Monitoring of High‐Rise Buildings Using Seismic Interferometry , 2017 .

[32]  Nori Nakata,et al.  Damage Detection of a Building Caused by the 2011 Tohoku-Oki Earthquake with Seismic Interferometry , 2015 .

[33]  Ah Chung Tsoi,et al.  Face recognition: a convolutional neural-network approach , 1997, IEEE Trans. Neural Networks.

[34]  H Y Kim,et al.  STATISTICAL ANALYSIS OF FRAGILITY CURVES , 2000 .

[35]  Rih-Teng Wu,et al.  Data fusion approaches for structural health monitoring and system identification: Past, present, and future , 2018, Structural Health Monitoring.

[36]  Andrew W. Smyth,et al.  Application of the unscented Kalman filter for real‐time nonlinear structural system identification , 2007 .

[37]  Rih-Teng Wu,et al.  Deep Convolutional Neural Network for Structural Dynamic Response Estimation and System Identification , 2019, Journal of Engineering Mechanics.

[39]  Mark J. van der Laan,et al.  A Method to Identify Significant Clusters in Gene Expression Data , 2002 .

[40]  Tianqi Chen,et al.  Empirical Evaluation of Rectified Activations in Convolutional Network , 2015, ArXiv.

[41]  C. M. Wen,et al.  A neural network approach for structural identification and diagnosis of a building from seismic response data , 2003 .

[42]  Lawrence G. Griffis,et al.  Serviceability Limit States Under Wind Load , 2003 .

[43]  James M. W. Brownjohn,et al.  Dynamic Assessment of Curved Cable-Stayed Bridge by Model Updating , 2000 .

[44]  Christopher D. Stephens,et al.  Report on progress at the Center for Engineering Strong Motion Data (CESMD) , 2012 .

[45]  I. Iervolino,et al.  Computer Aided Civil and Infrastructure Engineering , 2009 .

[46]  Raimondo Betti,et al.  A Hybrid Optimization Algorithm with Bayesian Inference for Probabilistic Model Updating , 2015, Comput. Aided Civ. Infrastructure Eng..

[47]  P. A. Fishwick Neural network models in simulation: a comparison with traditional modeling approaches , 1989, WSC '89.

[48]  Nick Gregor,et al.  NGA Project Strong-Motion Database , 2008 .

[49]  Si-Da Zhou,et al.  A Data-Driven Response Virtual Sensor Technique with Partial Vibration Measurements Using Convolutional Neural Network , 2017, Sensors.

[50]  Lambros S. Katafygiotis,et al.  Model updating using noisy response measurements without knowledge of the input spectrum , 2005 .

[51]  Hao Sun,et al.  Impact load identification for composite structures using Bayesian regularization and unscented Kalman filter , 2017 .

[52]  Danilo P. Mandic,et al.  Recurrent Neural Networks for Prediction: Learning Algorithms, Architectures and Stability , 2001 .

[53]  Roel Snieder,et al.  Monitoring a Building Using Deconvolution Interferometry. II: Ambient‐Vibration Analysis , 2014 .

[54]  Jiuchao Feng,et al.  Real-time nonlinear structural system identification via iterated unscented Kalman filter , 2012 .

[55]  Kurt Hornik,et al.  Approximation capabilities of multilayer feedforward networks , 1991, Neural Networks.

[56]  Lai Ming,et al.  Nonlinear structural response prediction based on support vector machines , 2008 .

[57]  Andrea Belleri,et al.  Damage assessment through structural identification of a three‐story large‐scale precast concrete structure , 2014 .

[58]  George W. Irwin,et al.  Multi-layer perceptron based modelling of nonlinear systems , 1996, Fuzzy Sets Syst..

[59]  Ali S. Hadi,et al.  Finding Groups in Data: An Introduction to Chster Analysis , 1991 .

[60]  K. Yuen Bayesian Methods for Structural Dynamics and Civil Engineering , 2010 .

[61]  Hao Sun,et al.  Bayesian characterization of buildings using seismic interferometry on ambient vibrations , 2017 .

[62]  Hoon Sohn,et al.  A review of structural health monitoring literature 1996-2001 , 2002 .

[63]  Oral Büyüköztürk,et al.  Probabilistic updating of building models using incomplete modal data , 2016 .

[64]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[65]  Luca Maria Gambardella,et al.  Proceedings of the Twenty-Second International Joint Conference on Artificial Intelligence Flexible, High Performance Convolutional Neural Networks for Image Classification , 2022 .

[66]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[67]  Yongming Liu,et al.  Aircraft dynamics simulation using a novel physics-based learning method , 2019, Aerospace Science and Technology.

[68]  Tianping Chen,et al.  Approximation capability to functions of several variables, nonlinear functionals and operators by radial basis function neural networks , 1993, Proceedings of 1993 International Conference on Neural Networks (IJCNN-93-Nagoya, Japan).

[69]  Masanobu Shinozuka,et al.  Nonlinear Static Procedure for Fragility Curve Development , 2000 .

[70]  Roel Snieder,et al.  Monitoring a Building Using Deconvolution Interferometry. I: Earthquake‐Data Analysis , 2013 .

[71]  Wang Ying,et al.  Artificial Neural Network Prediction for Seismic Response of Bridge Structure , 2009, 2009 International Conference on Artificial Intelligence and Computational Intelligence.

[72]  Shirley J. Dyke,et al.  Real-Time Dynamic Model Updating of a Hysteretic Structural System , 2014 .

[73]  Andreas Stavridis,et al.  System identification and modeling of a dynamically tested and gradually damaged 10‐story reinforced concrete building , 2018 .

[74]  Martín Abadi,et al.  TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems , 2016, ArXiv.