On continuous health monitoring of bridges under serious environmental variability by an innovative multi-task unsupervised learning method

Entezami, Alireza; Sarmadi, Hassan; Behkamal, Bahareh; Michele, Carlo De

doi:10.1080/15732479.2023.2166538

Cited by 20 publications

(6 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Once the structural features have been obtained, decision-making can be implemented by different machine learning algorithms. In this regard, one can exploit the main learning algorithms, in terms of supervised learning and unsupervised learning [9], and leverage some advanced algorithms, which include but are not limited to deep learning [10], transfer learning [11], kernel learning [12], ensemble learning [13], empirical learning [14], metalearning [15], multi-task learning [16], etc. Totally, these learning algorithms intend to develop computational models by using training data and then use the trained models for some tasks, such as classification, prediction, regression, clustering, and anomaly detection via test data.…”

Section: Introductionmentioning

confidence: 99%

Elimination of Thermal Effects from Limited Structural Displacements Based on Remote Sensing by Machine Learning Techniques

et al. 2023

Self Cite

View full text Add to dashboard Cite

Confounding variability caused by environmental and/or operational conditions is a big challenge in the structural health monitoring (SHM) of large-scale civil structures. The elimination of such variability is of paramount importance in avoiding economic and human losses. Machine learning-aided data normalization provides a good solution to this challenge. Despite proper studies on data normalization using structural responses/features acquired from contact-based sensors, this issue has not been explored properly via new features, such as displacement responses from remote sensing products, including synthetic aperture radar (SAR) images. Hence, the main aim of this work was to eliminate environmental variability, particularly thermal effects, from different and limited structural displacements retrieved from a few SAR images related to long-term health monitoring programs of long-span bridges. For this purpose, we conducted a comprehensive comparative study to investigate two supervised and two unsupervised data normalization algorithms. The supervised algorithms were based on Gaussian process regression (GPR) and support vector regression (SVR), for which temperature records acquired from contact temperature sensors and structural displacements retrieved from spaceborne remote sensors produce univariate predictor (input) and response (output) data for the regression problem. For the unsupervised algorithms, this paper employed principal component analysis (PCA) and proposed a deep autoencoder (DAE), both of which conform with unsupervised reconstruction-based data normalization. In contrast to the GPR- and SVR-based data normalization algorithms, both the PCA and DAE methods only consider the SAR-based displacement (output) data without any requirement of the environmental and/or operational (input) data. Limited displacement sets of long-span bridges from a few SAR images of Sentinel-1A, related to long-term SHM programs, were considered to assess the aforementioned techniques. Results demonstrate that the proposed DAE-aided data normalization is the best approach to remove thermal effects and other unmeasured environmental and/or operational variability.

show abstract

Section: Introductionmentioning

confidence: 99%

Elimination of Thermal Effects from Limited Structural Displacements Based on Remote Sensing by Machine Learning Techniques

et al. 2023

Self Cite

View full text Add to dashboard Cite

show abstract

“…4–6 Machine learning is the main area of artificial intelligence that intends to develop an automated learner (i.e., computational model) via training data and then conduct some tasks in terms of classification, regression, prediction, clustering, anomaly detection, etc. This methodology contains some underlying frameworks based on supervised, 7 semi-supervised, 8 and unsupervised learning 9 within some advanced algorithms under deep learning, 10 transfer learning, 11 active learning, 12 kernel learning, 13 multi-task learning, 14 dictionary learning, 15 meta-learning, 16 etc.…”

Section: Introductionmentioning

confidence: 99%

Unsupervised data normalization for continuous dynamic monitoring by an innovative hybrid feature weighting-selection algorithm and natural nearest neighbor searching

Sarmadi

Entezami

Magalhães

2023

Structural Health Monitoring

Self Cite

View full text Add to dashboard Cite

Continuous dynamic monitoring brings an important opportunity to evaluate the health and integrity of civil structures in a long-term manner. However, high dimensionality and sparsity of data caused by long-term monitoring and negative influences of environmental and/or operational variability are major challenges in this process. To address these important issues, this article proposes an innovative unsupervised data normalization method based on a novel hybrid feature weighting-selection algorithm and the idea of natural nearest neighbor (NN) searching emanated from the theory of mutual friendships in human societies. The proposed hybrid algorithm is a combination of global feature weighting with a new weighting measure and local feature selection. For this algorithm, this article leverages the natural NN searching that seeks to find adequate NNs automatically. The main objective of the proposed method is to remove the environmental and/or operational effects and provide normalized weighted features for reliable continuous dynamic monitoring. Using such features, an anomaly detector based on the Mahalanobis-squared distance is developed to assess and detect structural damage. The key innovations of this paper contain proposing a fully nonparametric unsupervised learning technique in two parts of data normalization and anomaly detection and developing a novel hybrid algorithm for removing the environmental and/or operational variations. Long-term dynamic features (modal frequencies) of a three-span box-girder concrete bridge (Z24 Bridge) and a long-span concrete arch bridge (Infante Dom Henrique Bridge) are considered to verify the proposed technique with several comparisons. Results indicate that this technique is successful and reliable in mitigating the environmental and/or operational effects and notifying accurate structural states.

show abstract

“…Based on the nature of the available training data, machine learning can generally be categorized into three classes: supervised, semisupervised, and unsupervised learning. The suitability of these algorithms for SHM applications was succinctly explored by Entezami et al [30]. Unsupervised learning, in particular, holds significant promise for a wide range of SHM applications, especially in the context of early damage assessment, as it does not necessitate fully labeled data, meaning there is no requirement for prior knowledge about the current or potential damage state of a structure.…”

Section: Introductionmentioning

confidence: 99%

Data Anomaly Detection for Structural Health Monitoring Based on a Convolutional Neural Network

Kim,

Mukhiddinov

2023

Sensors

View full text Add to dashboard Cite

Structural health monitoring (SHM) has been extensively utilized in civil infrastructures for several decades. The status of civil constructions is monitored in real time using a wide variety of sensors; however, determining the true state of a structure can be difficult due to the presence of abnormalities in the acquired data. Extreme weather, faulty sensors, and structural damage are common causes of these abnormalities. For civil structure monitoring to be successful, abnormalities must be detected quickly. In addition, one form of abnormality generally predominates the SHM data, which might be a problem for civil infrastructure data. The current state of anomaly detection is severely hampered by this imbalance. Even cutting-edge damage diagnostic methods are useless without proper data-cleansing processes. In order to solve this problem, this study suggests a hyper-parameter-tuned convolutional neural network (CNN) for multiclass unbalanced anomaly detection. A multiclass time series of anomaly data from a real-world cable-stayed bridge is used to test the 1D CNN model, and the dataset is balanced by supplementing the data as necessary. An overall accuracy of 97.6% was achieved by balancing the database using data augmentation to enlarge the dataset, as shown in the research.

show abstract

On continuous health monitoring of bridges under serious environmental variability by an innovative multi-task unsupervised learning method

Cited by 20 publications

References 45 publications

Elimination of Thermal Effects from Limited Structural Displacements Based on Remote Sensing by Machine Learning Techniques

Elimination of Thermal Effects from Limited Structural Displacements Based on Remote Sensing by Machine Learning Techniques

Unsupervised data normalization for continuous dynamic monitoring by an innovative hybrid feature weighting-selection algorithm and natural nearest neighbor searching

Data Anomaly Detection for Structural Health Monitoring Based on a Convolutional Neural Network

Contact Info

Product

Resources

About