Bridge damage identification using deep learning-based Convolutional Neural Networks (CNNs)

Mapping Intimacies ◽

10.31224/osf.io/ahtpm ◽

2021 ◽

Author(s):

Sandeep Sony

Keyword(s):

Neural Network ◽

Damage Identification ◽

Severe Damage ◽

Acceleration Time ◽

One Dimensional ◽

Damage Scenarios ◽

Benchmark Data ◽

Bridge Damage ◽

Novel Method ◽

Damage Types

In this paper, a novel method is proposed based on windowed-one dimensional convolutional neural network for multiclass damage detection using acceleration responses. The data is pre-processed and augmented by extracting samples of windows of the original acceleration time-series. 1D CNN is developed to classify the signals in multiple classes. The damage is detected if the predicted classification is one of the indicated damage levels. The damage is quantified using the predicted class probabilities. Various signals from the accelerometers are provided as input to 1D CNN model, and the resulting class probabilities are used to identify the location of the damage. The proposed method is validated using Z24 bridge benchmark data for multiclass classification for two damage scenarios. The results show that the proposed 1D CNN methods performs with superior accuracy for severe damage cases and works well with different type of damage types.

Download Full-text

Conventional Bridge Damage Identification Based on BP Neural Network

IOP Conference Series Materials Science and Engineering ◽

10.1088/1757-899x/730/1/012037 ◽

2020 ◽

Vol 730 ◽

pp. 012037

Author(s):

Zeying Yang ◽

Yalei Zhang ◽

Yaping Wang ◽

Ning Wang ◽

Haina Cui ◽

...

Keyword(s):

Neural Network ◽

Bp Neural Network ◽

Damage Identification ◽

Bridge Damage

Download Full-text

Damage Identification for Prestressed Adjacent Box-Beam Bridges

Advances in Civil Engineering ◽

10.1155/2014/540363 ◽

2014 ◽

Vol 2014 ◽

pp. 1-16 ◽

Cited By ~ 1

Author(s):

Kenneth K. Walsh ◽

Brendan T. Kelly ◽

Eric P. Steinberg

Keyword(s):

Damage Identification ◽

Longitudinal Direction ◽

Longitudinal Axis ◽

Element Model ◽

Modal Parameters ◽

Civil Infrastructure ◽

Response Data ◽

Acceleration Data ◽

Bridge Damage ◽

Damage Types

Structural health monitoring (SHM) has gained considerable attention as a tool for monitoring the health of civil infrastructure. For bridge infrastructure, previous methods have focused on the detection of localized damage through modal parameters extracted from the longitudinal direction of the structure. This paper investigates a new damage detection method based on the change in the first vertical mode extracted from the transverse direction of the bridge. The mode is determined through application of modal curve fitting to frequency response functions (FRFs) that are formed using vertical response data obtained in the direction perpendicular to the bridge’s longitudinal axis. Using this method, both local damage and global damage in the bridge reveal themselves as having a localized effect on the bridge response. Furthermore, damage is revealed in such a way that it enables differentiation of the damage types. To demonstrate the effectiveness of the method, modal parameters were extracted from acceleration data obtained from a finite element model of a full bridge. Analysis of the modal parameters showed that the proposed approach could not only detect both local and global bridge damage, but could also differentiate between damage types using only one mode shape. The proposed method was compared to a previously developed SHM method.

Download Full-text

Automatic Concrete Damage Recognition Using Multi-Level Attention Convolutional Neural Network

Materials ◽

10.3390/ma13235549 ◽

2020 ◽

Vol 13 (23) ◽

pp. 5549

Author(s):

Hyun Kyu Shin ◽

Yong Han Ahn ◽

Sang Hyo Lee ◽

Ha Young Kim

Keyword(s):

Neural Network ◽

Damage Detection ◽

Convolutional Neural Network ◽

Damage Identification ◽

Concrete Surface ◽

Superior Performance ◽

Detection Accuracy ◽

Damage Recognition ◽

Proposed Model ◽

Damage Types

There has been an increase in the deterioration of buildings and infrastructure in dense urban regions, and several defects in the structures are being exposed. To ensure the effective diagnosis of building conditions, vision-based automatic damage recognition techniques have been developed. However, conventional image processing techniques have some limitations in real-world situations owing to their manual feature extraction approach. To overcome these limitations, a convolutional neural network-based image recognition technique was adopted in this study, and a convolution-based concrete multi-damage recognition neural network (CMDnet) was developed. The image datasets consisted of 1981 types of concrete surface damages, including surface cracks, rebar exposure and delamination, as well as intact. Furthermore, it was experimentally demonstrated that the proposed model could accurately classify the damage types. The results obtained in this study reveal that the proposed model can recognize the different damage types from digital images of the surfaces of concrete structures. The trained CMDnet demonstrated a damage-detection accuracy of 98.9%. Moreover, the proposed model could be applied in automatic damage detection networks to achieve superior performance with regard to concrete surface damage detection and recognition, as well as accelerating efficient damage identification during the diagnosis of deteriorating structures used in civil engineering applications.

Download Full-text

SPATIAL ENTROPY OF ONE-DIMENSIONAL CELLULAR NEURAL NETWORK

International Journal of Bifurcation and Chaos ◽

10.1142/s0218127400001304 ◽

2000 ◽

Vol 10 (09) ◽

pp. 2129-2140 ◽

Cited By ~ 8

Author(s):

SONG-SUN LIN ◽

TZI-SHENG YANG

Keyword(s):

Neural Network ◽

Parameter Space ◽

Cellular Neural Network ◽

Mosaic Pattern ◽

Global Pattern ◽

One Dimensional ◽

Spatial Entropy ◽

Novel Method

This work investigates the global mosaic pattern and spatial entropy for one-dimensional cellular neural network (CNN). A novel method is developed to partition the parameter space into finitely many regions. The CNNs, with parameters in each region, have the same global pattern. An algorithm is also presented to evaluate the spatial entropy.

Download Full-text