Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2019 Sep 19;19(18):4035.
doi: 10.3390/s19184035.

A Machine Learning Approach to Bridge-Damage Detection Using Responses Measured on a Passing Vehicle

Abdollah Malekjafarian  1 , Fatemeh Golpayegani  2 , Callum Moloney  3 , Siobhán Clarke  4
Affiliations

A Machine Learning Approach to Bridge-Damage Detection Using Responses Measured on a Passing Vehicle

Abdollah Malekjafarian et al. Sensors (Basel). .

Abstract

This paper proposes a new two-stage machine learning approach for bridge damage detection using the responses measured on a passing vehicle. In the first stage, an artificial neural network (ANN) is trained using the vehicle responses measured from multiple passes (training data set) over a healthy bridge. The vehicle acceleration or Discrete Fourier Transform (DFT) spectrum of the acceleration is used. The vehicle response is predicted from its speed for multiple passes (monitoring data set) over the bridge. Root-mean-square error is used to calculate the prediction error, which indicates the differences between the predicted and measured responses for each passage. In the second stage of the proposed method, a damage indicator is defined using a Gaussian process that detects the changes in the distribution of the prediction errors. It is suggested that if the bridge condition is healthy, the distribution of the prediction errors will remain low. A recognizable change in the distribution might indicate a damage in the bridge. The performance of the proposed approach was evaluated using numerical case studies of vehicle-bridge interaction. It was demonstrated that the approach could successfully detect the damage in the presence of road roughness profile and measurement noise, even for low damage levels.

Keywords: artificial neural network; bridge; damage detection; drive-by; machine learning.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Figure 1
Figure 1
The proposed algorithm.
Figure 2
Figure 2
Neurons output calculation.
Figure 3
Figure 3
Finite element (FE) model of a quarter-car passing over a bridge.
Figure 4
Figure 4
The response measured on the axle of the vehicle passing over healthy and damaged bridges (x/L = vt/L) (a) acceleration and (b) FFT spectrum. Reproduced from OBrien et al. Application of empirical mode decomposition to drive-by bridge damage detection. Eur. J. Mech. A Solid 2017; 61, 151–163. Copyright © 2017 Elsevier Masson SAS. All rights reserved [25].
Figure 4
Figure 4
The response measured on the axle of the vehicle passing over healthy and damaged bridges (x/L = vt/L) (a) acceleration and (b) FFT spectrum. Reproduced from OBrien et al. Application of empirical mode decomposition to drive-by bridge damage detection. Eur. J. Mech. A Solid 2017; 61, 151–163. Copyright © 2017 Elsevier Masson SAS. All rights reserved [25].
Figure 5
Figure 5
The comparison of the predicted and measured vehicle responses for the first four passes over the healthy bridge.
Figure 6
Figure 6
The comparison of the predicted and measured vehicle responses for the first four passes over the damaged bridge with crack ratio of 0.3.
Figure 7
Figure 7
(a) The prediction errors and (b) DI for the healthy and damaged cases (DI: damage indicator).
Figure 8
Figure 8
Two quarter-cars passing over the bridge.
Figure 9
Figure 9
The prediction errors for (a) 1% added noise, (b) 3% added noise, and (c) 5% added noise (dark blue: healthy, red: 5% damage, yellow: 10% damage, purple: 15% damage, green: 20% damage, light blue: 25% damage, brown: 30% damage).
Figure 10
Figure 10
The DIs for (a) 1% added noise; (b) 3% added noise; and (c) 5% added noise.
Figure 11
Figure 11
The comparison of the predicted and measured vehicle FFT responses over the healthy bridge.
Figure 12
Figure 12
The comparison of the predicted and measured vehicle FFT responses over the damaged bridge with a crack ratio of 0.3.
Figure 13
Figure 13
(a) The prediction errors and (b) the DI for the healthy and damaged bridges.
Figure 14
Figure 14
The prediction errors for (a) 1% added noise, (b) 3% added noise, and (c) 5% added noise.
Figure 15
Figure 15
The DIs for (a) 1% added noise, (b) 3% added noise, and (c) 5% added noise.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Malekjafarian A., OBrien E.J., Golpayegani F. Indirect monitoring of critical transport infrastructure: Data analysis and signal processing. In: Alavi A., Buttlar W.G., editors. Data Anlytics Applications for Smart Cities. Auerbach/CRC Press; Boca Raton, FL, USA: 2018.
    1. Malekjafarian A., McGetrick P.J., OBrien E.J. A Review of Indirect Bridge Monitoring Using Passing Vehicles. Shock Vib. 2015;2015:1–16. doi: 10.1155/2015/286139. - DOI
    1. Yang Y.B., Lin C.W., Yau J.D. Extracting bridge frequencies from the dynamic response of a passing vehicle. J. Sound Vib. 2004;272:471–493. doi: 10.1016/S0022-460X(03)00378-X. - DOI
    1. Kim C.W., Kawatani M. Challenge for a drive-by bridge inspection; Proceedings of the 10th International Conference on Structural Safety and Reliability ICOSSAR2009; Osaka, Japan. 13–17 September 2009; pp. 758–765.
    1. Yang Y.B., Chang K.C. Extracting the bridge frequencies indirectly from a passing vehicle: Parametric study. Eng. Struct. 2009;31:2448–2459. doi: 10.1016/j.engstruct.2009.06.001. - DOI

LinkOut - more resources

  NODES
INTERN 1
Project 1
twitter 2