Deep autoencoder architecture for bridge damage assessment using responses from several vehicles

Sarwar, Muhammad Zohaib; Cantero, Daniel

doi:10.1016/j.engstruct.2021.113064

Cited by 64 publications

(21 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The tool VEqMon2D has been already used in recent publications, see [11,12]. These numerical studies required the use of more elaborated vehicle models with different body configurations and axle arrangements.…”

Section: Impactmentioning

confidence: 99%

VEqMon2D—Equations of motion generation tool of 2D vehicles with Matlab

Cantero

2022

SoftwareX

Self Cite

View full text Add to dashboard Cite

Section: Impactmentioning

confidence: 99%

VEqMon2D—Equations of motion generation tool of 2D vehicles with Matlab

Cantero

2022

SoftwareX

Self Cite

View full text Add to dashboard Cite

“…( 8), 𝑴 𝒗 , 𝑪 𝒗 , and 𝑲 𝒗 are the mass, damping and stiffness matrices, while 𝑢 𝑣 contains the displacements of all DOFs of the vehicle model. The vehicle parameters and their variability are taken from [12], for the realisation of Monte Carlo simulations. The values of the vehicle parameters are mainly based on European 5-axles trucks and adopted from [38,39].…”

Section: Numerical Modellingmentioning

confidence: 99%

“…In order to study the effect of noise, white Gaussian noise is added to the acceleration signals by using Eq. (12).…”

Section: Effect Of Measurement Noisementioning

confidence: 99%

“…The reported methods mainly rely on advanced signal processing techniques and machine learning methods for damage detection. For instance, [11,12] used data driven techniques and statistical analysis for damage detection and quantification. While [13,14] estimate the contact point response between vehicle and bridge for damage detection and localisation.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Vehicle Assisted Bridge Damage Assessment Using Probabilistic Deep Learning

Sarwar¹,

Cantero²

2022

SSRN Journal

View full text Add to dashboard Cite

“…Liu et al [ 31 ] achieved damage quantification of bridges by utilizing stacked autoencoders as dimensionality reduction techniques to explore the vehicle’s full bandwidth frequency response. Sarwar and Cantero [ 32 ] effectively detected bridge damage by using a fleet of vehicles’ responses and adopting deep autoencoders as feature extraction techniques. These methods, however, rely on ML techniques used as signal processing tools for feature extraction.…”

Section: Introductionmentioning

confidence: 99%

A Time-Domain Signal Processing Algorithm for Data-Driven Drive-by Inspection Methods: An Experimental Study

Lan

Lin

2023

Materials

View full text Add to dashboard Cite

Constructional material deterioration and member damage can cause changes in the dynamic characteristics of bridge structures, and such changes can be tracked in the responses of passing vehicles via the vehicle-bridge interaction (VBI). Though data-driven methods have shown promising results in damage inspection for drive-by methods, there is still much room for improvement in their performance. Given this background, this paper proposes a novel time-domain signal processing algorithm for the raw vehicle acceleration data of data-driven drive-by inspection methods. To achieve the best data processing performance, an optimizing strategy is designed to automatically search for the optimal parameters, tuning the algorithm. The proposed method intentionally overcomes the difficulties in the application of drive-by methods, such as measurement noise, speed variance, and enormous data volumes. Meanwhile, the use of this method can greatly improve the accuracy and efficiency of Machine Learning (ML) models in vehicle-based damage detection. It consists of a filtering process to denoise the data, a pooling process to reduce data redundancy, and an optimizing procedure to maximize algorithm performance. A dataset is obtained to validate the proposed algorithm through laboratory experiments with a scale truck model and a steel beam. The results show that, compared to using raw data, the present algorithm can increase the average accuracy by 12.2–15.0%, and the average efficiency by 35.7–96.7% for different damaged cases and ML models. Additionally, the functions of filtering and pooling operations, the influence of window function parameters, as well as the performance of different sensor locations, are also investigated in the paper. The goal is to present a signal processing algorithm for data-driven drive-by inspection methods to improve their detection performance of bridge damage caused by material deterioration or structural change.

show abstract

Deep autoencoder architecture for bridge damage assessment using responses from several vehicles

Cited by 64 publications

References 53 publications

VEqMon2D—Equations of motion generation tool of 2D vehicles with Matlab

VEqMon2D—Equations of motion generation tool of 2D vehicles with Matlab

Vehicle Assisted Bridge Damage Assessment Using Probabilistic Deep Learning

A Time-Domain Signal Processing Algorithm for Data-Driven Drive-by Inspection Methods: An Experimental Study

Contact Info

Product

Resources

About