Online unsupervised detection of structural changes using train–induced dynamic responses

Meixedo, Andreia; Santos, João; Ribeiro, Diogo; Calçada, Rui; Todd, Michael D.

doi:10.1016/j.ymssp.2021.108268

Cited by 54 publications

(42 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Lastly, in ongoing and future work, using the validated train–track–bridge dynamic model, the authors intend to develop an advanced methodology for structural damage identification based on artificial intelligence [ 54 , 55 ].…”

Section: Discussionmentioning

confidence: 99%

Train–Track–Bridge Dynamic Interaction on a Bowstring-Arch Railway Bridge: Advanced Modeling and Experimental Validation

Ribeiro

Calçada

Brehm³

et al. 2022

Sensors

Self Cite

View full text Add to dashboard Cite

This article describes the validation of a 3D dynamic interaction model of the train–track–bridge system on a bowstring-arch railway bridge based on experimental tests. The train, track, and bridge subsystems were modeled on the basis of large-scale and highly complex finite elements models previously calibrated on the basis of experimental modal parameters. The train–bridge dynamic interaction problem, in the vertical direction, was efficiently solved using a dedicated computational application (TBI software). This software resorts to an uncoupled methodology that considers the two subsystems, bridge and train, as two independent structures and uses an iterative procedure to guarantee the compatibility of the forces and displacements at the contact points at each timestep. The bridge subsystem is solved by the mode superposition method, while the train subsystem is solved by a direct integration method. The track irregularities were included in the dynamic problem based on real measurements performed by a track inspection vehicle. A dynamic test under traffic actions allowed measuring the responses in the bridge, track, and vehicles, which were synchronized by GPS systems. The test results demonstrated the occurrence of upward displacements on the deck, which is a characteristic of structures with an arch structural behavior, as well as an alternation of tensile/compressive stresses between the rail and deck due to the deck–track composite effect. Furthermore, the acceleration response of the bridge proved to be significantly influenced by the train operating speed. The validation procedure involved comparing the dynamic responses obtained from the train–bridge interaction model, including track irregularities, and the responses obtained experimentally, through the test under traffic actions. A very good correlation was obtained between numerical and experimental results in terms of accelerations, displacements, and strains. The contributions derived from the parametric excitation of the train, the global/local dynamic behavior of the bridge, and the excitation derived from the track irregularities were decisive to accurately reproduce the complex behavior of the train–track–bridge system.

show abstract

Section: Discussionmentioning

confidence: 99%

Train–Track–Bridge Dynamic Interaction on a Bowstring-Arch Railway Bridge: Advanced Modeling and Experimental Validation

Ribeiro

Calçada

Brehm³

et al. 2022

Sensors

Self Cite

View full text Add to dashboard Cite

show abstract

“…Previous studies demonstrate good results in railway defect detection using these approaches, namely, in the detection of train wheel damages, such as flats [ 26 , 27 , 28 , 29 , 30 , 31 ], out-of-roundness [ 32 ], and squats and corrugation [ 33 ]. Typically, these damage identification techniques require several operations including [ 34 , 35 ]: (i) data acquisition, (ii) feature extraction, (iii) feature normalization, (iv) feature fusion, and (v) feature classification. Other innovative techniques for wheel flat fault detection include a time–frequency ridge estimation method [ 36 ] or a multiscale morphology analysis [ 29 ] that can effectively extract the influential features of signals from strong background noise and under variable speed conditions.…”

Section: Introductionmentioning

confidence: 99%

“…Symbolic data, continuous wavelet transform [ 27 , 37 ], principal component analysis (PCA) [ 38 , 39 ], and autoregressive models [ 34 , 40 ] are examples of effective techniques for extracting damage-sensitive features for both static and dynamic monitoring. In applications where the measured quantity is acceleration, autoregressive models (AR) have been widely disseminated, and the autoregressive model with exogenous input (ARX) [ 35 ] shows a higher sensitivity to damage in relation to AR due to its ability to capture cross information between sensors.…”

Section: Introductionmentioning

confidence: 99%

“…The final operation of the damage detection methodology is feature classification. Typically, two different approaches are used: (i) unsupervised methods, in which models are trained using labelled data under the supervision of training data, such as k-means [ 34 ], self-organizing maps (SOM) [ 47 ], and cluster analysis [ 35 ], or (ii) supervised methods, in which models are not supervised using training dataset, such as Naive Bayes classifiers [ 48 ] and k-Nearest Neighbour classifier [ 49 ].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Early Identification of Unbalanced Freight Traffic Loads Based on Wayside Monitoring and Artificial Intelligence

Silva

Guedes

Ribeiro

et al. 2023

Sensors

View full text Add to dashboard Cite

The identification of instability problems in freight trains circulation such as unbalanced loads is of particular importance for railways management companies and operators. The early detection of unbalanced loads prevents significant damages that may cause service interruptions or derailments with high financial costs. This study aims to develop a methodology capable of automatically identifying unbalanced vertical loads considering the limits proposed by the reference guidelines. The research relies on a 3D numerical simulation of the train–track dynamic response to the presence of longitudinal and transverse scenarios of unbalanced vertical loads and resorting to a virtual wayside monitoring system. This methodology is based on measured data from accelerometers and strain gauges installed on the rail and involves the following steps: (i) feature extraction, (ii) features normalization based on a latent variable method, (iii) data fusion, and (iv) feature discrimination based on an outlier and a cluster analysis. Regarding feature extraction, the performance of ARX and PCA models is compared. The results prove that the methodology is able to accurately detect and classify longitudinal and transverse unbalanced loads with a reduced number of sensors.

show abstract

“…Besides, different methods and models such as the application of modal reduction and mode superposition technique for efficient estimation of critical (highly stressed) areas in civil structures [24], spatial incompatibility filters [25], data mining [26], [27], artificial neural network [28], clustering analysis [8], genetic algorithm [29], deep learning [30], fuzzy logic [31], principal component analysis [32], Bayesian [33], support vector machine [34], particle swarm optimization [35], decision tree [36], regression analysis [37], remote sensing [38], unmanned aerial systems [39], [40], and anomaly detection [41] have been also applied in SHM.…”

Section: Introductionmentioning

confidence: 99%

Implementation of a Secure Storage Using Blockchain for PCA-FRF Sensor Data of Plate-Like Structures

et al. 2022

View full text Add to dashboard Cite

Structural Health Monitoring (SHM) systems have widely been used to guarantee the safe functioning of electrical, mechanical, civil, and aerospace engineering assets. It is required to improve the SHM systems from different aspects in terms of smartification, performance, beneficiation, sustainment, automation, cost-effectivity, and safety using cutting-edge technologies. Blockchain is currently the most revolutionary technology in computer science. Implementation of Blockchain in different fields of academia and industry has recently given very good results. However, application of Blockchain in SHM is still in the infancy stage. Therefore, many challenges are still ahead. One of the main potential applications of Blockchain is to secure the sensor data. In this study, a local Blockchain approach is developed as a secure storage using the extracted Principal Components (PCs) of Frequency Response Function (FRF) data obtained from modal analysis of a plate-like structure. In general, Secure Hash Algorithm (SHA) is one of the most practical hash functions with efficient performance which has been employed in Bitcoin. Therefore, in this research, SHA-256 is considered to generate the hash for each block. To the best of our knowledge, this is the first attempt to develop a secure storage for SHM data using Blockchain.

show abstract

Online unsupervised detection of structural changes using train–induced dynamic responses

Cited by 54 publications

References 35 publications

Train–Track–Bridge Dynamic Interaction on a Bowstring-Arch Railway Bridge: Advanced Modeling and Experimental Validation

Train–Track–Bridge Dynamic Interaction on a Bowstring-Arch Railway Bridge: Advanced Modeling and Experimental Validation

Early Identification of Unbalanced Freight Traffic Loads Based on Wayside Monitoring and Artificial Intelligence

Implementation of a Secure Storage Using Blockchain for PCA-FRF Sensor Data of Plate-Like Structures

Contact Info

Product

Resources

About