Faster Post-Earthquake Damage Assessment Based on 1D Convolutional Neural Networks

Abstract

Contemporary deep learning approaches for post-earthquake damage assessments based on 2D convolutional neural networks (CNNs) require encoding of ground motion records to transform their inherent 1D time series to 2D images, thus requiring high computing time and resources. This study develops a 1D CNN model to avoid the costly 2D image encoding. The 1D CNN model is compared with a 2D CNN model with wavelet transform encoding and a feedforward neural network (FNN) model to evaluate prediction performance and computational efficiency. A case study of a benchmark reinforced concrete (r/c) building indicated that the 1D CNN model achieved a prediction accuracy of 81.0%, which was very close to the 81.6% prediction accuracy of the 2D CNN model and much higher than the 70.8% prediction accuracy of the FNN model. At the same time, the 1D CNN model reduced computing time by more than 90% and reduced resources used by more than 69%, as compared to the 2D CNN model. Therefore, the developed 1D CNN model is recommended for rapid and accurate resultant damage assessment after earthquakes.

Special Issue on Application of Artificial Neural Networks for Seismic Design and Assessment

“…The inherent ability of ANNs to incorporate the knowledge gained from previous procedures (through re-training), as well as their ability for the instant extraction of results, are the two main characteristics which give the potential for a rapid estimation of the SDS of numerous of buildings in stricken areas. This fact-which was investigated in several previous studies-is reflected in the research papers of Harirchian et al [3] and Yuan et al [4].…”

“…Yuan et al [4] studied three different types of neural network models (1D Convolutional Neural Networks, 1D-CNN; 2D Convolutional Neural Networks, 2D-CNN; and Feedforward Neural Network, FNN) as regards their ability to effectively predict the SDS of RC buildings in near-real time. This evaluation was based on a benchmark RC frame building, while 1993 historical ground excitations properly selected from the PEER Ground Motion Records (GMR) database were used.…”

Data‐driven probabilistic curvature capacity modeling of circular RC columns facilitating seismic fragility analyses of highway bridges

“…ANN, a bioinspired classical machine learning approach, has been one of the most prevalent tools for solving structural engineering problems. [17][18][19][20] As illustrated in Figure 1, an ANN is commonly composed of (a) an input layer with multiple input nodes, each representing a specific parameter of the assessed structure or associated external loads such as earthquake excitations, (b) multiple hidden layers, each containing multiple hidden neurons (or called hidden nodes), and (c) an output layer with multiple output nodes that refer to structural responses under the external loads. More specifically, the ith hidden node in the first hidden layer, h i [1] , can be calculated through:…”

Fusing damage-sensitive features and domain adaptation towards robust damage classification in real buildings

“…ML algorithms provide powerful tools for training classifiers to differentiate healthy from damaged monitoring data or even attribute data to various classes related to increasing severity of damage [30][31][32][33][34]. In the context of post-earthquake assessment, Yuan et al [35] trained a convolutional neural network (NN) for post-earthquake damage prediction, reaching an accuracy of 80% in model predictions. However, generalizing predictions from ML algorithms, trained on simulated data, to real structures is not trivial, while such references are very limited in the current literature.…”