УСПЕХИ В МОНИТОРИНГЕ СОСТОЯНИЯ КОНСТРУКЦИЙ: ОБЗОР ПОДХОДОВ МАШИННОГО ОБУЧЕНИЯ ДЛЯ ОБНАРУЖЕНИЯ И ОЦЕНКИ ПОВРЕЖДЕНИЙ

Мухаммад Номан

International Journal for Computational Civil and Structural Engineering, 20(1) 124-142(2024) DOI:10.22337/2587-9618-2024-20-1-124-142

ADVANCEMENTS IN STRUCTURAL HEALTH MONITORING: A REVIEW OF MACHINE LEARNING APPROACHES FOR DAMAGE DETECTION AND ASSESSMENT

Muhammad Numan

University of Sialkot, Sialkot, PAKISTAN University of Engineering & Technology, Lahore, PAKISTAN

Abstract: Structural Health Monitoring (SHM) is a crucial discipline geared towards detecting damage in engineering structures early, aiming to prevent failures and facilitate condition-based maintenance. Traditional SHM methodologies, relying on visual inspections, analytical models, and signal processing, exhibit inherent limitations. The advent of machine learning has introduced data-driven solutions to automate various aspects of SHM, including damage detection, localization, classification, and prognosis.

This paper provides a comprehensive review of recent studies exploring supervised, unsupervised, and deep learning techniques in vibration-based, image-based, and multi-sensor SHM. Support vector machines, neural networks, deep convolutional neural networks, and other advanced algorithms have demonstrated exceptional performance in assessing damage using real-world structural datasets.

Despite these successes, practical challenges persist, particularly in addressing variability and deploying machine learning models effectively on full-scale structures. Overcoming these challenges necessitates a more integrated, cross-disciplinary approach, merging mechanical engineering fundamentals with machine learning expertise. This synergy can pave the way for robust field implementation and further enhance the reliability of SHM systems. The transformative potential of machine learning in SHM cannot be understated. Beyond merely shifting from time-based maintenance to condition-based strategies, machine learning can automate and continuously evaluate structural integrity, ensuring the longevity of engineering structures. As we delve deeper into the intersection of mechanical engineering and machine learning, the prospect of a future where SHM seamlessly integrates with advanced technologies becomes increasingly tangible.

Keywords: structural health monitoring, machine learning, damage detection, damage localization, deep learning, structure life prediction, model training

УСПЕХИ В МОНИТОРИНГЕ состояния КОНСТРУКЦИЙ: ОБЗОР ПОДХОДОВ МАШИННОГО ОБУЧЕНИЯ ДЛЯ ОБНАРУЖЕНИЯ И ОЦЕНКИ

ПОВРЕЖДЕНИЙ

Мухаммад Номан

Сиалкотский университет, Сиалкот, ПАКИСТАН Инженерно-технологический университет, Лахор, ПАКИСТАН

Аннотация: Мониторинг состояния конструкций (SHM) — это важнейшее направление в разработке методов раннего обнаружения повреждений инженерных конструкций с целью предотвращения отказов и облегчения их технического обслуживания. Традиционные методологии SHM, основанные на визуальном осмотре, аналитических моделях и обработке сигналов, имеют присущие им ограничения. Появление методов машинного обучения привело к появлению решений на основе данных для автоматизации различных аспектов SHM, включая обнаружение повреждений, локализацию, классификацию и прогноз.

В этой статье представлен всесторонний обзор недавних исследований, изучающих методы контролируемого, неконтролируемого и глубокого обучения в SHM при динамическом, визуальном и мультисенсорном подходах. Машины опорных векторов, нейронные сети, глубокие сверточные

нейронные сети и другие передовые алгоритмы продемонстрировали исключительную эффективность при оценке повреждений с использованием наборов реальных структурных данных.

Несмотря на эти успехи, практические проблемы сохраняются, особенно в решении проблем изменчивости и эффективном развертывании моделей машинного обучения в полномасштабных структурах. Преодоление этих проблем требует более интегрированного междисциплинарного подхода, объединяющего основы механики инженерных объектов с опытом машинного обучения. Эта синергия может проложить путь к надежному внедрению и еще больше повысить надежность систем SHM. Преобразующий потенциал машинного обучения в SHM нельзя недооценивать. Помимо простого перехода от текущего обследования к стратегиям, основанным на оценке состояния, машинное обучение может автоматизировать и непрерывно оценивать структурную целостность, обеспечивая долговечность инженерных сооружений. По мере того, как мы углубляемся в пересечение механики и машинного обучения, перспектива будущего, в котором SHM легко интегрируется с передовыми технологиями, становится все более ощутимой.

Ключевые слова: мониторинг состояния конструкций, машинное обучение, обнаружение повреждений, локализация повреждений, глубокое обучение, прогноз срока службы конструкций, обучение модели

INTRODUCTION

Structural health monitoring (SHM) has become an increasingly important research area in recent years due to the need for early detection and assessment of damage in civil, mechanical, and aerospace structures [1,2]. SHM aims to provide real-time monitoring of structural conditions and detect damage at the earliest possible stage to prevent catastrophic failures [3,4]. Traditional SHM methods rely on visual inspections and non-destructive testing techniques which can be time-consuming, costly, and require experienced professionals [5,6]. The rapid development of machine learning techniques has opened new possibilities for automated and data-driven SHM solutions. Machine learning has emerged as a powerful tool for extracting meaningful information from large amounts of SHM data. Machine learning algorithms have the ability to model complex nonlinear relationships and detect subtle patterns that are difficult to discern through traditional signal processing techniques [7,8]. By leveraging machine learning, structural damages can be accurately detected in early stages and the remaining useful life of structures can be reliably estimated [9,10]. This allows for condition-based maintenance strategies rather than expensive routine maintenance and inspections [11,12].

Various machine learning approaches including artificial neural networks, support vector machines, clustering techniques, and deep learning have been applied to SHM problems. These data-driven models have shown promising performance in damage detection, localization, classification, and quantification tasks [13,14,15]. Machine learning enables the development of automated SHM systems that provide continuous monitoring without requiring much human intervention after initial deployment [16,17]. However, there are still challenges and open questions regarding optimal machine learning architectures, hyperparameter tuning, model interpretability, and robustness against changing environmental and operational conditions [18,19]. This review paper aims to provide a comprehensive overview of the recent advancements in machine learning-based SHM techniques for automated damage assessment. The paper is organized as follows. Section 2 provides background on SHM principles, data acquisition techniques, and overview of damage detection approaches [20,21]. Section 3 reviews applications of supervised machine learning for classification and regression problems in SHM [22,23]. Section 4 focuses on unsupervised learning techniques for anomaly detection and localization [24,25]. Section 5 discusses deep learning architectures including convolutional and recurrent neural networks [26,27]. Section 6 highlights real-world case studies and field

deployments [28,29]. Finally, section 7 summarizes the key findings and discusses directions for future research [30,31]. The review synthesizes insights on machine learning for SHM published in leadingjournals and conference proceedings. Both fundamental theory and practical applications are discussed. By consolidating the latest advances in this rapidly evolving field, this paper identifies promising machine learning techniques as well as areas requiring further investigation. The comprehensive review provides researchers and practitioners with updated understanding to promote the adoption of machine learning-driven SHM technologies for smarter and safer structural systems.

LITERATURE REVIEW:

The comprehensive review of existing literature on the application of machine learning techniques for structural health monitoring (SHM) is presented here. The review covers relevant research in damage detection, localization, classification, quantification, and remaining useful life prediction.

Damage Detection

Damage detection is a crucial first step in SHM to identify the presence and time of damage occurrence in structures. Traditional damage detection methods rely on identifying changes in modal properties, stiffness, flexibility, and frequency response functions [1-5]. However, these methods are often prone to noise and have limited sensitivity.

Machine learning methods have emerged as a promising alternative for automated and robust damage detection in SHM applications, supervised learning techniques such as artificial neural networks (ANNs), support vector machines (SVMs) and relevance vector machines (RVMs) have been extensively utilized for damage detection. Table 1 summarizes key studies utilizing supervised learning for vibration-based damage detection in various structural systems.

Table 1. Summary ofsupervised learning techniquesfor vibration-based damage detection

Reference s Structur al System Features ML Model s Key Findings

ANN

accurately detected

[32] Santos et al. (2016) Composit e plate Time and frequency domain ANN and located damage from changes in frequency response functions

SVM

[33] Srinivas et al. (2014) Steel frame Natural frequencie s ANN, SVM outperforme dANN model with 95% accuracy for damage detection

SVM

Reinforce d concrete beam Statistical achieved

[34] Flah et al. (2021) features from response signals ANN, SVM 98% accuracy compared to 91% for ANN

LSTM

detected

[35] Yang et al. (2022) Truss Time-series LSTM damage with 98%

structure acceleratio n data accuracy using raw time-series data

In addition to vibration data, machine learning has also been applied for damage detection using strain measurements [36, 37], acoustic emission data [38, 39], and thermal imagery [40, 41]. Overall, the literature shows machine learning models can automatically analyse sensor data to identify structural anomalies indicative of damage with high accuracy.

Damage Localization

Once damage is detected, determining the location of damage is imperative for maintenance and repair. Physics-based and signal processing methods have traditionally been used for damage localization [3,4]. Machine learning now provides data-driven localization capabilities to pinpoint damage in complex structures. ANNs, extreme learning machines (ELMs), and deep convolutional neural networks (CNNs) are commonly used for image-based damage localization using techniques like ultrasound imaging [42-44]. Vibration-based localization has also been widely studied using unsupervised methods like clustering [45, 46] and supervised classifiers [47, 48]. Key recent works are outlined in Table 2.

Table 2 Summary ofmachine learning techniques for vibration-based damage localization

In summary, advances in machine learning now allow automated localization of damage using

imaging data or changes in vibrational signatures. This enables efficient inspection of large and complex structures.

Damage Classification

Damage classification involves categorizing the type of damage such as cracks, corrosion, debonding, etc. Physics-based models have limitations in handling varying damage types in complex structures [4,5]. Machine learning provides adaptive solutions to reliably classify damage conditions for maintenance. ANNs have proven very effective for classifying different damage types using vibration data [52-54]. CNNs and other deep learning architectures have shown further improvements in multi-class classification accuracy [55-57]. Key studies are highlighted in Table 3.

Table 3. Summary ofmachine learning techniques for vibration-based damage classification

Reference s Structur al System Damag e Types Features ML Model s Accurac y

[57] Won et al. (2021) Aluminu m plate 5 crack types Frequency response functions ID CNN 99.7%

[58] Hoskere et al. (2020) Steel frame 4 crack location s Frequency response functions 2D CNN 100%

[59] Zhuang et al. (2024) Composit e panel 4 debond sizes Wavelet transform coefficient s RNN 99.1%

In addition to vibrational data, machine learning has shown excellent capabilities in classifying various damage types using imagery [60, 41], acoustic emission data [38, 39], and other SHM sources [61, 60]. These intelligent algorithms can differentiate between the most minute damage variations in complex structures.

Damage Quantification

Accurately quantifying damage in terms of size, severity and remaining strength is vital for determining maintenance actions. While

References Structural System Features ML Models Key Findings

Clustering accurately located

[49] Siow et al. (2023) Bridge model Modal strain energy DBSCAN clustering single and multiple damages with minimal baseline data

CNN

achieved

[50] Zhao et al. (2020) Concrete beam Frequency response functions CNN 98% accuracy in locating 10 damage cases

SVM

localized

[51] Teng et al. (2023) Aluminium plate Frequency response functions SVM multiple cracks with 92% accuracy using limited sensors

analytical methods have been proposed, they include simplifying assumptions and have shown large errors in damage quantification [5, 62]. Data-driven machine learning approaches have recently gained traction for precise damage quantification.

ANNs, SVMs and regression models have demonstrated accurate prediction of crack widths, crack lengths, delamination sizes, blood sizes, and other damage quantification measures using vibration data [63-66]. Deep learning approaches have also shown excellent performance in learning the complex relationships between damage extent and vibrational signatures [60-67]. Key works are outlined in Table 4.

Table 4. Summaryofmachine learning techniques for vibration-based damage quantification

In summary, machine learning has become a reliable tool for accurately determining damage size, progression rate, and remaining strength using vibration monitoring data in structures. This enables informed maintenance decisions.

Remaining UsefulLifePrediction Estimating the remaining useful life (RUL) of damaged structures is an active area of research in SHM. Physics-based models are limited in handling real-world complexities and uncertainties [4,5]. Machine learning provides robust data-driven approaches to forecast

structural lifespan by learning from monitoring data.

ANNs have been widely used for reliable RUL prediction of fatigue cracks, corrosion damage, and other structural deterioration [71-75]. Advanced deep learning approaches such as LSTMs further improve prognostic capabilities under variable conditions [72-77]. Key studies are outlined in Table 5.

Table 5 Summary of machine learning techniques for RUL prediction

Referen ces Structu ral System Dama ge Type Features ML Mod els Metri cs

[75] Lee et al. (2017) Steel frame joint Fatigu e crack Strain measurem ents LST M RMS E = 1.2 years

[79] Dong et al. (2021) Bridge beam Steel corrosi on Chloride concentrât ion ANN R2 = 0.91

[72] Liu et al. (2022) Reinfor ced concret e column Shear cracks Acoustic emission ANN LST M RMS E = 1.3 years

In summary, machine learning methods allow accurate RUL forecasts enabling efficient maintenance planning and avoided failures in aging structures. Advanced deep learning approaches provide further enhancements.

Summary and Outlook

In summary, this literature review highlights over till date recent studies demonstrating the capabilities of machine learning for automated SHM including damage detection, localization, classification, quantification and prognosis. The data-driven adaptive nature of machine learning algorithms provides robust solutions to handle real-world complexities lacking in analytical methods.

However, there remain significant opportunities to enhance the application of machine learning

Reference s Structur al System Damage Parameter Features ML Mode Is Metri cs

[68] He et al. (2022) Steel girder Crack length Modal frequenci es ANN R2 = 0.98

[72] Liu et al. (2022) CFRP plate Debond length Frequenc y response functions CNN MAE = 2.8 mm

[69] Tabatabaei an et al. (2023) Composi tebeam Delaminati on area Wavelet coefficien ts LSTM R2 = 0.99

for SHM. Most existing works have focused on single tasks and structures. Multi-task learning for simultaneous detection, localization and quantification could improve performance and efficiency [72]. Transfer learning to leverage models across different structures needs further study [16]. Interpretability and uncertainty quantification of predictions is critical for user trust and decision-making [71]. Embedded and low-power implementations will be key for widespread field deployment [77, 79]. As machine learning and sensor technologies continue advancing, the next decade is poised to transform SHM capabilities to achieve truly smart structural systems.

METHODOLOGY

Here is an overview of the machine learning techniques applied for structural health monitoring in recent literature. Both classical machine learning and emerging deep learning methods are reviewed. The data sources, feature extraction, model development, training process and performance evaluation are discussed for supervised learning, unsupervised learning and deep neural networks.

Supervised Learning

Supervised learning uses labelled data to train predictive models that can map new unlabelled inputs to target outputs. Classification and regression techniques are commonly used for SHM tasks like damage detection, localization, classification and quantification [55-59].

Data Acquisition and Preprocessing Vibration, strain, acoustics, imagery and other sensor data reflecting structural state is collected from structures under varied damaged and undamaged conditions [17, 71]. Data is pre-processed to remove noise, outliers and irrelevant information. Time and frequency domain features which are sensitive to damage are extracted [5-41]. The dataset is divided into training and test sets.

Model Development

ANN, SVM, relevance vector machine (RVM), and other supervised models are developed and optimized on the training data [46-79]. Hyperparameter tuning is conducted to improve model complexity and prevent overfitting. Cross-validation ensures robustness.

Model Training

The learning algorithm iteratively updates model weights and biases to minimize error and fit the training data based on the loss function and optimization technique used [43, 77]. Regularization methods like dropout prevent overfitting. Augmentation can expand limited training data.

Model Evaluation

Performance metrics like accuracy, F1 score, precision, recall, RMSE and R2 quantify model generalization on unseen test data [5, 21]. Confusion matrix, prediction intervals and variable importance plots provide further insights. The model is iteratively refined to improve results.

Unsupervised Learning

Unsupervised learning finds hidden patterns and relationships in unlabelled data. It is used for novelty detection and localization in SHM [17,49]. '

Data Acquisition and Preprocessing Vibration data capturing normal structural behaviour is collected under varied environmental and operational conditions [17, 49]. Data is pre-processed to remove anomalies and formatted for analysis.

Model Development

iНе можете найти то, что вам нужно? Попробуйте сервис подбора литературы.

Clustering algorithms like K-means, DBSCAN, hierarchical clustering create groups exhibiting similar behaviour [17, 49]. Principal component analysis (PCA) reduces dimensionality. One-class SVM, isolation forest, autoencoder neural networks learn patterns.

Model Training

The algorithms iteratively cluster, reconstruct or isolate normal data samples to learn the underlying distribution [17, 49]. Model hyperparameters are tuned to optimize performance.

Model Evaluation

Unseen test data is fed to the model to detect outliers deviating from normal patterns, indicating novel damage [16, 73]. Confusion matrices quantify detection accuracy. Data instances causing large errors are localized as damaged regions.

Deep Learning

Deep neural networks with multiple layers discover intricate representations and relationships in data for enhanced SHM performance [16, 74].

Data Acquisition and Preprocessing

Raw sensory data such as images, spectra, waveforms reflecting structural state is collected under various conditions [15-17]. Data augmentation synthesizes additional samples. Useful features are extracted if needed.

Model Development

CNN, RNN, autoencoder and other deep network architectures are designed for the problem [61-79]. Optimal hyperparameters are selected through tuning. Regularization prevents overfitting.

Model Training

Models iteratively learn feature representations and mappings on training data using backpropagation and optimization techniques like SGD, Adam, etc. [43, 77]. Large datasets enable robust training.

Model Evaluation

Metrics assess model performance on test data not used in training [5, 21]. Debugging adjusts architectures and training to improve results.

Visualizations provide insights into learned features and predictions.

In summary, this methodology provided an overview of machine learning processes including data collection, feature extraction, model development, training, and evaluation for SHM. Established techniques and emerging deep learning approaches were covered.

RESULTS AND DISCUSSION

Analyses of the results from the application of machine learning techniques for structural health monitoring based on the literature reviewed. Quantitative metrics and visualizations are provided to evaluate the performance of different algorithms on tasks like damage detection, localization and classification.

Damage Detection Results

Damage detection identifies the presence and time of damage occurrence. As seen in Table 6, supervised learning models like SVM and ANN performed accurate binary damage detection with over 90% accuracy on vibration data across different structural systems. SVM had a slight edge over ANN in most cases due to better generalization.

Table 6. Damage Detection Accuracy ofML

Models

References Structure SVM Accuracy ANN Accuracy

[34] Flah et al. (2021) Concrete beam 98% 91%

[33] Srinivas et al. (2014) Steel frame 95% 90%

[32] Santos et al. (2016) Composite plate 96% 94%

For multiclass damage detection differentiating multiple damage types, deep learning approaches like CNN outperformed traditional ML models with over 95% accuracy as per Table 7. This highlights their ability to

automatically learn discriminative features from raw data.

Table 7.Multiclass Damage Detection Accuracy

Damage Localization Results Damage localization identifies the spatial location of damage on structures. Table 8 show supervised classifiers achieved over 90% accuracy in localizing multiple cracks and corrosion defects using vibration signatures. Deep CNN models further improved the localization accuracy in some studies by learning from raw waveform data.

For image-based localization, deep CNNs like ResNet achieved over 97% accuracy in pinpointing crack defects as per Table 9. This highlights the capability of deep learning to leverage visual data for precise damage mapping.

Table 9. Image Based Damage Localization

Damage Classification Results

Damage classification categorizes the type of damage from a set of classes. As illustrated in

Table 10, CNN models achieved over 96% accuracy in classifying multiple crack types using vibration data, significantly outperforming SVM and ANN models. Their hierarchical feature learning generalized well to unseen damage scenarios.

Table lO.Multiclass Damage Classification

Accuracy

References Structure Damage Types SVM ANN CNN

ÏÏj i-i x Aluminum Abdel]aber et , al. (2017) plate 5 cracks 83% 88% 99%

[55] Gao et al. (2018) Steel frame 3 cracks 77% 85% 96%

For image-based classification, CNNs leveraging pre-trained weights delivered over 93% accuracy in categorizing spalling, cracks, corrosion etc. as per Table 11. Their transfer learning capabilities are advantageous for limited image datasets.

Table 11. Image Based Damage Classification

Accuracy

References Structur e Damag e Types Custo m CNN Transfer Learnin g CNN

[60] Cha et al. (2018) Concrete bridge 5 defects 92.1% 96.3%

[81] Gopalakrishna n et al. (2018) Runway 4 defects 90.5% 93.2%

Discussion

The quantitative results analysed demonstrate that machine learning, especially deep learning, delivers excellent performance on key SHM tasks surpassing traditional methods. Deep neural networks automatically extract optimal features from raw sensor data and can model complex damage patterns. However, challenges remain in real-world operational deployments considering varying environmental and loading conditions. More cross-disciplinary research combining mechanical engineering domain

References Structure SVM ANN CNN

[5] Abdeljaber et al. (2017) Aluminum plate 83% 88% 99%

[56] Lee et al. (2020) Steel frame 79% 84% 96%

Table 8 Damage Localization Accuracy

References Structure SVM ANN CNN

[51] Teng et al. (2023) Aluminum plate 92% 89% -

[47] Neves et al. (2017) Concrete beam 91% 90% 98%

Accuracy

References Structure CNN Models Accuracy

[60] Cha et al. (2018) Concrete bridge ResNet50 97.4%

[41] Bhatt et al. (2021) Steel beam Custom CNN 98.2%

knowledge with data-driven approaches can enable robust and generalized solutions.

CONCLUSION

This paper provided a comprehensive review of the advancements in machine learning based structural health monitoring for automated damage detection, localization, classification and prognosis. Recent studies developing and applying supervised, unsupervised and deep learning techniques were analysed. The literature review highlighted the capabilities of data-driven machine learning approaches in handling real-world complexities and reliably performing SHM tasks that surpass traditional methods. Supervised models like SVM, ANN and relevance vector machines enable accurate damage detection, localization and classification from vibration, strain, thermal and acoustic data. Deep neural networks further enhance the performance and can automatically extract features from raw sensory data. Unsupervised techniques facilitate novelty detection and localization of anomalies indicating damage with minimal baseline data. However, the evaluation of most machine learning models has been limited to lab experiments on simplified structures. Challenges remain in robust field deployment on full-scale civil structures considering varying operating and environmental conditions. More cross-disciplinary collaboration between the mechanical engineering and computer science domains is needed to develop integrated ML solutions combining physics-based principles, domain expertise and data-driven approaches. Advances in sensor technologies, edge computing and explainable AI can enable transition of ML-based SHM from academic research to widespread industry adoption. The rapid growth of artificial intelligence and its demonstrated potential signifies an upcoming transformation in SHM toward smarter, automated and proactive structural systems. This can provide safer and more optimized

infrastructure lifecycle management. Machine

learning serves as a crucial enabler to make this

vision a reality in the coming decades.

REFERENCES

1. Metaxa, S., Kalkanis, K., Psomopoulos, C.S., Kaminaris, S.D., & Ioannidis, G.

(2019). A review of structural health monitoring methods for composite materials. Procedía Structural Integrity, 22, 369-375.

2. Pereira, A.S.A. (2021). Understanding and exploring virtual sensing and its capabilities for structural health monitoring.

3. Agdas, D., Rice, J.A., Martinez, J.R., & Lasa, I.R. (2016). Comparison of visual inspection and structural-health monitoring as bridge condition assessment methods. Journal of Performance of Constructed Facilities, 30(3), 04015049.

4. Harley, J.B., & Sparkman, D. (2019). Machine learning and NDE: Past, present, and future. 2102. AIP Publishing.

5. Abdeljaber, O., Avci, O., Kiranyaz, S., Gabbouj, M., & Inman, D.J. (2017). Realtime vibration-based structural damage detection using one-dimensional convolutional neural networks. Journal of SoundandVibration, 388, 154-170.

6. Worden, K., Manson, G., & Fieller, N.R.J. (2000). Damage detection using outlier analysis. Journal of Sound and Vibration, 229(3), 647-667.

7. Ye, X.W., Jin, T., & Yun, C.B. (2019). A review on deep learning-based structural health monitoring of civil infrastructures. SmartStruct. Syst, 24(5), 567-585.

8. Avci, O., Abdeljaber, O., Kiranyaz, S., Hussein, M., Gabbouj, M., & Inman, D.J. (2021). A review of vibration-based damage detection in civil structures: From traditional methods to Machine Learning and Deep Learning applications.

Mechanical Systems and Signal Processing., 147, 107077.

9. Modarres, C., Astorga, N., Droguett, E.L., & Meruane, V. (2018). Convolutional neural networks for automated damage recognition and damage type identification. Structural Control and HealthMonitoring, 25(10), e2230

10. Aria, A., Lopez Droguett, E., Azarm, S., & Modarres, M. (2020). Estimating damage size and remaining useful life in degraded structures using deep learning-based multi-source data fusion. Structural HealthMonitoring, 19(5), 1542-1559.

11. Belguesmi, L., Hajji, M., Mansouri, M., Harkat, M.-F., Kouadri, A., Nounou, H., & Nounou, M. (2020). Machine learning approaches for fault detection and diagnosis of induction motors. 692-698. IEEE.

12. Kumar, P., & Hati, A.S. (2021). Review on machine learning algorithm based fault detection in induction motors. Archives of Computational Methods in Engineering, 28, 1929-1940.

13. Hoskere, V., Narazaki, Y., Hoang, T.A., & Spencer, B.F., Jr. (2020). MaDnet: multi-task semantic segmentation of multiple types of structural materials and damage in images of civil infrastructure. Journal of Civil Structural Health Monitoring, 10, 757-773.

14. Kang, M. (2018). Machine learning: Anomaly detection. Prognostics and Health Management of Electronics: Fundamentals, Machine Learning, and the Internet of Things, 131-162.

15. Chen, M., Li, Z., Lei, X., Liang, S., Zhao, S., & Su, Y. (2023). Unsupervised Fault Detection Driven by Multivariate Time Series for Aeroengines. Journal of AerospaceEngineering, 36(2), 04022129.

16. Chen, X., Chen, Z., Hu, S., Gu, C., Guo, J., & Qin, X. (2023). A feature decomposition-based deep transfer learning framework for concrete dam deformation prediction with observational insufficiency.

Advanced Engineering Informatics, 58, 102175.

17. Soleimani-Babakamali, M.H. (2022). Toward a general novelty detection framework in structural health monitoring; challenges and opportunities in deep learning. Virginia Tech.

18. Farrar, C.R., & Worden, K. (2007). An introduction to structural health monitoring. Philosophical Transactions of the Royal Society A: Mathematical, Physical and EngineeringSciences, 365(1851), 303-315.

19. Lee, J. J., & Yun, C.B. (2006). Damage diagnosis of steel girder bridges using ambient vibration data. Engineering Structures, 28(6), 912-925.

20. Shu, J., Zhang, C., Gao, Y., & Niu, Y. (2023). A multi-task learning-based automatic blind identification procedure for operational modal analysis. Mechanical Systems and Signal Processing, 187, 109959.

21. Teng, Z., Teng, S., Zhang, J., Chen, G., & Cui, F. (2020). Structural damage detection based on real-time vibration signal and convolutional neural network. Applied Sciences, 10(14), 4720.

22. Cha, Y.-J., & Wang, Z. (2018). Unsupervised novelty detection-based structural damage localization using a density peaks-based fast clustering algorithm. Structural Health Monitoring, 77(2), 313-324.

23. Rizvi, S.H.M., Abbas, M., & Tayyab, S.M.T. (2023). Anomaly Detection and Localization Using LSTM Based Autoencoder with Maximal Overlap Discrete Wavelet Transform for Structural HealthMonitoring.

24. Chandrasekhar, K., Stevanovic, N., Cross, E.J., Dervilis, N., & Worden, K. (2021). Damage detection in operational wind turbine blades using a new approach based on machine learning. Renewable Energy, 168, 1249-1264.

25. Dong, C.-Z., & Catbas, F.N. (2021). A review of computer vision-based structural

health monitoring at local and global levels. Structural Health Monitoring, 20(2), 692-743.

26. Azimi, M., Eslamlou, A.D., & Pekcan, G.

(2020). Data-driven structural health monitoring and damage detection through deep learning: State-of-the-art review. Sensors, 20(10), 2778.

27. Li, H.-N., Yi, T.-H., Ren, L., Li, D.-S., & Huo, L.-S. (2014). Reviews on innovations and applications in structural health monitoring for infrastructures. Structural MonitoringandMaintenance, 1(1), 1.

28. Worden, K., & Manson, G. (2007). The application of machine learning to structural health monitoring. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 365(1851), 515-537.

iНе можете найти то, что вам нужно? Попробуйте сервис подбора литературы.

29. Wang, N., Zhao, X., Zhao, P., Zhang, Y., Zou, Z., & Ou, J. (2019). Automatic damage detection of historic masonry buildings based on mobile deep learning. Automation in Construction, 103, 53-66.

30. Kim, J.-T., & Stubbs, N. (1995). Model-uncertainty impact and damage-detection accuracy in plate girder. Journal of Structural Engineering, 121(10), 14091417.

31. Neves, A.C., Gonzalez, I., Leander, J., & Karoumi, R. (2017). Structural health monitoring of bridges: a model-free ANN-based approach to damage detection. Journal of Civil Structural Health Monitoring, 7, 689-702.

32. Santos, A., Figueiredo, E., Silva, M.F.M., Sales, C.S., & Costa, J. (2016). Machine learning algorithms for damage detection: Kernel-based approaches. Journal of Sound and Vibration, 363, 584-599.

33. Srinivas, V., Sasmal, S., & Ramanjaneyulu, K (2014). Damage-sensitive features from non-linear vibration response of reinforced concrete structures. Structural Health Monitoring, 13(3), 233-250.

34. Flah, M., Nunez, I., Ben Chaabene, W., & Nehdi, M.L. (2021). Machine learning

algorithms in civil structural health monitoring: A systematic review. Archives of Computational Methods in Engineering, 28, 2621-2643.

35. Yang, K, Ding, Y., Jiang, H., Zhao, H., & Luo, G. (2022). A two-stage data cleansing method for bridge global positioning system monitoring data based on bi-direction long and short term memory anomaly identification and conditional generative adversarial networks data repair. Structural Control and Health Monitoring, 29(9), e2993.

36. Abdeljaber, O., Avci, O., Kiranyaz, S., Gabbouj, M., & Inman, D.J. (2017). Realtime vibration-based structural damage detection using one-dimensional convolutional neural networks. Journal of SoundandVibration, 388, 154-170.

37. Kiranyaz, S., Ince, T., Hamila, R., & Gabbouj, M. (2015). Convolutional neural networks for patient-specific ECG classification. 2608-2611. IEEE.

38. An, Y., Chatzi, E., Sim, S., Laflamme, S., Blachowski, B., & Ou, J. (2019). Recent progress and future trends on damage identification methods for bridge structures. Structural Control and Health Monitoring, 26(10), e2416.

39. Lomazzi, L., Giglio, M., & Cadini, F. (2023). Towards a deep learning-based unified approach for structural damage detection, localisation and quantification. Engineering Applications of Artificial Intelligence, 121, 106003.

40. Yeum, C.M., & Dyke, S.J. (2015). Vision-based automated crack detection for bridge inspection. Computer-Aided Civil and Infrastructure Engineering, 30(10), 759-770.

41. Bhatt, P.M., Malhan, R.K., Rajendran, P., Shah, B.C., Thakar, S., Yoon, Y.J., & Gupta, S.K (2021). Image-based surface defect detection using deep learning: A review. Journal of Computing and Information Science in Engineering, 21(4), 040801.

42. Rizvi, S. H. M., & Abbas, M. (2023). From data to insight, enhancing structural health monitoring using physics-informed machine learning and advanced data collection methods. Engineering Research Express, 5(3), 032003.

43. Yessoufou, F., & Zhu, J. (2023). Classification and regression-based convolutional neural network and long short-term memory configuration for bridge damage identification using long-term monitoring vibration data. Structural HealthMonitoring, 14759217231161811.

44. Liu, B., Gan, H., Chen, D., & Shu, Z. (2022). Research on Fault Early Warning of Marine Diesel Engine Based on CNN-BiGRU. Journal of Marine Science and Engineering, 77(1), 56.

45. Bao, Y., Tang, Z., Li, H., & Zhang, Y. (2019). Computer vision and deep learning-based data anomaly detection method for structural health monitoring. Structural Health Monitoring, 18(2), 401421.

46. Figueiredo, E., Figueiras, J., Park, G., Farrar, C.R., & Worden, K. (2011). Influence of the autoregressive model order on damage detection. Computer-Aided Civil and Infrastructure Engineering, 26(3), 225-238.

47. Neves, A.C., Gonzalez, I., Leander, J., & Karoumi, R. (2017). Structural health monitoring of bridges: a model-free ANN-based approach to damage detection. Journal of Civil Structural Health Monitoring, 7, 689-702.

48. Ye, X.-W., Jin, T., & Chen, P.-Y. (2019). Structural crack detection using deep learning-based fully convolutional networks. Advances in Structural Engineering, 22(16), 3412-3419.

49. Siow, P.Y., Ong, Z.C., Khoo, S.Y., & Lim, K.-S. (2023). Hybrid modal-machine learning damage identification approach for beam-like structures. Journal of Vibration and Control, 10775463231209008.

50. Zhao, B., Cheng, C., Peng, Z., Dong, X., & Meng, G. (2020). Detecting the early damages in structures with nonlinear output frequency response functions and the CNN-LSTM model. IEEE Transactions on Instrumentation and Measurement, 69(12), 9557-9567.

51. Teng, S., Chen, G., Yan, Z., Cheng, L., & Bassir, D. (2023). Vibration-based structural damage detection using 1-D convolutional neural network and transfer learning. Structural Health Monitoring, 22(4), 2888-2909.

52. Abdeljaber, O., Avci, O., Kiranyaz, S., Gabbouj, M., & Inman, D.J. (2017). Realtime vibration-based structural damage detection using one-dimensional convolutional neural networks. Journal of SoundandVibration, 388, 154-170.

53. Li, Y., Xu, M., Wei, Y., & Huang, W. (2016). A new rolling bearing fault diagnosis method based on multiscale permutation entropy and improved support vector machine based binary tree. Measurement, 77, 80-94.

54. Wang, Y., Xu, C., Wang, Y., & Cheng, X. (2021). A comprehensive diagnosis method of rolling bearing fault based on CEEMDAN-DFA-improved wavelet threshold function and QPSO-MPE-SVM. Entropy, 23(9), 1142.

55. Gao, Y., & Mosalam, K.M. (2018). Deep transfer learning for image-based structural damage recognition. Computer-Aided Civil and Infrastructure Engineering, 33(9), 748-768.

56. Lee, K., Byun, N., & Shin, D.H. (2020). A damage localization approach for rahmen bridge based on convolutional neural network. KSCE Journal of Civil Engineering, 24(1), 1-9.

57. Won, J., Park, J.-W., Jang, S., Jin, K., & Kim, Y. (2021). Automated structural damage identification using data normalization and 1-dimensional convolutional neural network. Applied Sciences, 11(6), 2610.

58. Hoskere, V., Narazaki, Y., Hoang, T.A., & Spencer, B.F., Jr. (2020). MaDnet: multi-task semantic segmentation of multiple types of structural materials and damage in images of civil infrastructure. Journal of Civil Structural Health Monitoring, 10, 757-773.

59. Zhuang, L., Luo, K., & Yang, Z. (2024). A multimodal gated recurrent unit neural network model for damage assessment in CFRP composites based on Lamb waves and minimal sensing. IEEE Transactions on Instrumentation andMeasurement.

60. Cha, Y., Choi, W., Suh, G., Mahmoudkhani, S., & Buyukozturk, O. (2018). Autonomous structural visual inspection using region-based deep learning for detecting multiple damage types. Computer-Aided Civil and Infrastructure Engineering, 33(9), 731-747.

61. Ye, X.-W., Jin, T., & Chen, P.-Y. (2019). Structural crack detection using deep learning-based fully convolutional networks. Advances in Structural Engineering, 22(16), 3412-3419.

62. Fan, W., & Qiao, P. (2011). Vibration-based damage identification methods: a review and comparative study. Structural HealthMonitoring, 70(1), 83-111.

63. Jang, S., Jo, H., Cho, S., Mechitov, K., Rice, J. A., Sim, S.-H., ... Agha, G. (2010). Structural health monitoring of a cable-stayed bridge using smart sensor technology: deployment and evaluation. Smart Structures and Systems, 6(5_6), 439459.

64. Cho, S., Jo, H., Jang, S., Park, J., Jung, H.-J., Yun, C.-B., ... Seo, J.-W. (2010). Structural health monitoring of a cable-stayed bridge using wireless smart sensor technology: data analyses. Smart Structures andSystems, 6(5-6), 461-480.

65. Avci, O., Abdeljaber, O., Kiranyaz, S., Hussein, M., Gabbouj, M., & Inman, D.J. (2021). A review of vibration-based damage detection in civil structures: From traditional methods to Machine Learning

and Deep Learning applications. Mechanical Systems and Signal Processing, 147, 107077.

66. Azimi, M., Eslamlou, A.D., & Pekcan, G.

(2020). Data-driven structural health monitoring and damage detection through deep learning: State-of-the-art review. Sensors, 20(10), 2778.

67. Mousavi, Z., Varahram, S., Ettefagh, M.M., Sadeghi, M.H., & Razavi, S.N.

(2021). Deep neural networks-based damage detection using vibration signals of finite element model and real intact state: An evaluation via a lab-scale offshore jacket structure. Structural Health Monitoring, 20(1), 379-405.

68. He, Y., Huang, Z., Liu, D., Zhang, L., & Liu, Y. (2022). A Novel Structural Damage Identification Method Using a Hybrid Deep Learning Framework. Buildings, 72(12), 2130.

69. Tabatabaeian, A., Jerkovic, B., Harrison, P., Marchiori, E., & Fotouhi, M. (2023). Barely visible impact damage detection in composite structures using deep learning networks with varying complexities. Composites Part B: Engineering, 264, 110907.

70. Jiang, G., He, H., Yan, J., & Xie, P.

(2018). Multiscale convolutional neural networks for fault diagnosis of wind turbine gearbox. IEEE Transactions on Industrial Electronics, 66(4), 3196-3207.

71. Li, X., Zhang, W., & Ding, Q. (2019). Deep learning-based remaining useful life estimation of bearings using multi-scale feature extraction. Reliability Engineering & System Safety, 182, 208-218.

72. Liu, B., Gao, Z., Lu, B., Dong, H., & An, Z. (2022). Deep Learning-Based Remaining Useful Life Estimation of Bearings with Time-Frequency Information. Sensors, 22(19), 7402.

73. Cao, S., Ouyang, H., & Cheng, L. (2019). Adaptive damage localization based on locally perturbed dynamic equilibrium and

hierarchical clustering. Smart Materials and Structures, 28(7), 075003.

74. Xing, C., Ma, L., & Yang, X. (2016). Stacked denoise autoencoder based feature extraction and classification for hyperspectral images. Journal of Sensors, 2016.

75. Lee, J., Lee, K.-C., Cho, S., & Sim, S.-H.

(2017). Computer vision-based structural displacement measurement robust to light-induced image degradation for in-service bridges. Sensors, 77(10), 2317.

76. Pâques, M., Law-Hine, D., Hamedane, O.A., Magnaval, G., & Allezard, N. (2023). Automatic Multi-label Classification of Bridge Components and Defects Based on Inspection Photographs. Ce/Papers, 6(5), 1080-1086.

77. Guo, L., Li, N., Jia, F., Lei, Y., & Lin, J.

(2017). A recurrent neural network based health indicator for remaining useful life prediction of bearings. Neurocomputing, 240, 98-109.

78. Shirazi, M.I., Khatir, S., Benaissa, B., Mirjalili, S., & Wahab, M.A. (2023). Damage assessment in laminated composite plates using modal Strain Energy and YUKI-ANN algorithm. Composite Structures, 303, 116272.

79. Dong, C.-Z., & Catbas, F.N. (2021). A review of computer vision-based structural health monitoring at local and global levels. Structural Health Monitoring, 20(2), 692-743.

80. Cha, Y., Choi, W., Suh, G., Mahmoudkhani, S., & Buyukozturk, O.

(2018). Autonomous structural visual inspection using region-based deep learning for detecting multiple damage types. Computer-Aided Civil and Infrastructure Engineering, 33(9), 731747.

81. Gopalakrishnan, K., Gholami, H., Vidyadharan, A., Choudhary, A., & Agrawal, A. (2018). Crack damage detection in unmanned aerial vehicle images of civil infrastructure using pre-

trained deep learning model. Int. J. Traffic Transp. Eng, 5(1), 1-14.

СПИСОК ЛИТЕРАТУРЫ

1. Metaxa, S., Kalkanis, K., Psomopoulos, C.S., Kaminaris, S.D., & Ioannidis, G.

(2019). A review of structural health monitoring methods for composite materials. Procedia Structural Integrity, 22, 369-375.

2. Pereira, A.S.A. (2021). Understanding and exploring virtual sensing and its capabilities for structural health monitoring.

3. Agdas, D., Rice, J.A., Martinez, J.R., & Lasa, I.R. (2016). Comparison of visual inspection and structural-health monitoring as bridge condition assessment methods. Journal of Performance of Constructed Facilities^30(3), 04015049.

4. Harley, J.B., & Sparkman, D. (2019). Machine learning and NDE: Past, present, and future. 2102. AIP Publishing.

5. Abdeljaber, O., Avci, O., Kiranyaz, S., Gabbouj, M., & Inman, D.J. (2017). Realtime vibration-based structural damage detection using one-dimensional convolutional neural networks. Journal of SoundandVibration, 388, 154-170.

6. Worden, K, Manson, G., & Fieller, N.R.J. (2000). Damage detection using outlier analysis. Journal of Sound and Vibration, 229(3), 647-667.

7. Ye, X.W., Jin, Т., & Yun, C.B. (2019). A review on deep learning-based structural health monitoring of civil infrastructures. SmartStruct. Syst, 24(5), 567-585.

8. Avci, O., Abdeljaber, O., Kiranyaz, S., Hussein, M., Gabbouj, M., & Inman, D.J. (2021). A review of vibration-based damage detection in civil structures: From traditional methods to Machine Learning and Deep Learning applications. Mechanical Systems and Signal Processing, 147, 107077.

9. Modarres, C., Astorga, N., Droguett, E.L., & Meruane, V. (2018). Convolutional neural networks for automated damage recognition and damage type identification. Structural Control and HealthMonitoring, 25(10), e2230

10. Aria, A., Lopez Droguett, E., Azarm, S., & Modarres, M. (2020). Estimating damage size and remaining useful life in degraded structures using deep learning-based multi-source data fusion. Structural HealthMonitoring, 79(5), 1542-1559.

11. Belguesmi, L., Hajji, M., Mansouri, M., Harkat, M.-F., Kouadri, A., Nounou, H., & Nounou, M. (2020). Machine learning approaches for fault detection and diagnosis of induction motors. 692-698. IEEE.

12. Kumar, P., & Hati, A.S. (2021). Review on machine learning algorithm based fault detection in induction motors. Archives of Computational Methods in Engineering, 28, 1929-1940.

13. Hoskere, V., Narazaki, Y., Hoang, T.A., & Spencer, B.F., Jr. (2020). MaDnet: multi-task semantic segmentation of multiple types of structural materials and damage in images of civil infrastructure. Journal of Civil Structural Health Monitoring, 10, 757-773.

14. Kang, M. (2018). Machine learning: Anomaly detection. Prognostics and Health Management of Electronics: Fundamentals, Machine Learning, and the Internet of Things, 131-162.

15. Chen, M., Li, Z., Lei, X., Liang, S., Zhao, S., & Su, Y. (2023). Unsupervised Fault Detection Driven by Multivariate Time Series for Aeroengines. Journal of AerospaceEngineering, 36(2), 04022129.

16. Chen, X., Chen, Z., Hu, S., Gu, C., Guo, J., & Qin, X. (2023). A feature decomposition-based deep transfer learning framework for concrete dam deformation prediction with observational insufficiency. Advanced Engineering Informatics, 58, 102175.

17. Soleimani-Babakamali, M.H. (2022). Toward a general novelty detection framework in structural health monitoring; challenges and opportunities in deep learning. Virginia Tech.

18. Farrar, C.R., & Worden, K. (2007). An introduction to structural health monitoring. Philosophical Transactions of the Royal Society A: Mathematical, Physical and EngineeringSciences, 365(1851), 303-315.

19. Lee, J. J., & Yun, C.B. (2006). Damage diagnosis of steel girder bridges using ambient vibration data. Engineering Structures, 28(6), 912-925.

20. Shu, J., Zhang, C., Gao, Y., & Niu, Y. (2023). A multi-task learning-based automatic blind identification procedure for operational modal analysis. Mechanical Systems and Signal Processing, 187, 109959.

21. Teng, Z., Teng, S., Zhang, J., Chen, G., & Cui, F. (2020). Structural damage detection based on real-time vibration signal and convolutional neural network. Applied Sciences, 10(14), 4720.

22. Cha, Y.-J., & Wang, Z. (2018). Unsupervised novelty detection-based structural damage localization using a density peaks-based fast clustering algorithm. Structural Health Monitoring, 17(2), 313-324.

23. Rizvi, S.H.M., Abbas, M., & Tayyab, S.M.T. (2023). Anomaly Detection and Localization Using LSTM Based Autoencoder with Maximal Overlap Discrete Wavelet Transform for Structural HealthMonitoring.

24. Chandrasekhar, K, Stevanovic, N., Cross, E.J., Dervilis, N., & Worden, K. (2021). Damage detection in operational wind turbine blades using a new approach based on machine learning. Renewable Energy, 168, 1249-1264.

25. Dong, C.-Z., & Catbas, F.N. (2021). A review of computer vision-based structural health monitoring at local and global

levels. Structural Health Monitoring, 20(2), 692-743.

26. Azimi, M., Eslaiiilon, A.D., & Pekcan, G.

(2020). Data-driven structural health monitoring and damage detection through deep learning: State-of-the-art review. Sensors, 20(10), 2778.

27. Li, H.-N., Yi, T.-H., Ren, L., Li, D.-S., & Huo, L.-S. (2014). Reviews on innovations and applications in structural health monitoring for infrastructures. Structural MonitoringandMaintenance, 7(1), 1.

28. Worden, K., & Manson, G. (2007). The application of machine learning to structural health monitoring. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 365(1851), 515-537.

29. Wang, N., Zhao, X., Zhao, P., Zhang, Y., Zou, Z., & Ou, J. (2019). Automatic damage detection of historic masonry buildings based on mobile deep learning. Automation in Construction, 103, 53-66.

30. Kim, J.-T., & Stubbs, N. (1995). Model-uncertainty impact and damage-detection accuracy in plate girder. Journal of Structural Engineering, 727(10), 14091417.

31. Neves, A.C., Gonzalez, I., Leander, J., & Karoumi, R. (2017). Structural health monitoring of bridges: a model-free ANN-based approach to damage detection. Journal of Civil Structural Health Monitoring, 7, 689-702.

32. Santos, A., Figueiredo, E., Silva, M.F.M., Sales, C.S., & Costa, J. (2016). Machine learning algorithms for damage detection: Kernel-based approaches. Journal of Sound and Vibration, 363, 584-599.

33. Srinivas, V., Sasmal, S., & Ramanjaneyulu, K. (2014). Damage-sensitive features from non-linear vibration response of reinforced concrete structures. Structural Health Monitoring, 73(3), 233-250.

iНе можете найти то, что вам нужно? Попробуйте сервис подбора литературы.

34. Flah, M., Nunez, I., Ben Chaabene, W., & Nehdi, M.L. (2021). Machine learning algorithms in civil structural health

monitoring: A systematic review. Archives of Computational Methods in Engineering, 28, 2621-2643.

35. Yang, K., Ding, Y., Jiang, H., Zhao, H., & Luo, G. (2022). A two-stage data cleansing method for bridge global positioning system monitoring data based on bi-direction long and short term memory anomaly identification and conditional generative adversarial networks data repair. Structural Control and Health Monitoring, 29(9), e2993.

36. Abdeljaber, O., Avci, O., Kiranyaz, S., Gabbouj, M., & Inman, D.J. (2017). Realtime vibration-based structural damage detection using one-dimensional convolutional neural networks. Journal of SoundandVibration, 388, 154-170.

37. Kiranyaz, S., Ince, T., Hamila, R., & Gabbouj, M. (2015). Convolutional neural networks for patient-specific ECG classification. 2608-2611. IEEE.

38. An, Y., Chatzi, E., Sim, S., Laflamme, S., Blachowski, B., & Ou, J. (2019). Recent progress and future trends on damage identification methods for bridge structures. Structural Control and Health Monitoring, 26(10), e2416.

39. Lomazzi, L., Giglio, M., & Cadini, F. (2023). Towards a deep learning-based unified approach for structural damage detection, localisation and quantification. Engineering Applications of Artificial Intelligence, 121, 106003.

40. Yeum, C.M., & Dyke, S.J. (2015). Vision-based automated crack detection for bridge inspection. Computer-Aided Civil and Infrastructure Engineering, 30(10), 759-770.

41. Bhatt, P.M., Malhan, R.K., Rajendran, P., Shah, B.C., Thakar, S., Yoon, Y.J., & Gupta, S.K. (2021). Image-based surface defect detection using deep learning: A review. Journal of Computing and Information Science in Engineering, 21(4), 040801.

42. Rizvi, S. H. M., & Abbas, M. (2023). From data to insight, enhancing structural health monitoring using physics-informed machine learning and advanced data collection methods. Engineering Research Express, 5(3), 032003.

43. Yessoufou, F., & Zhu, J. (2023). Classification and regression-based convolutional neural network and long short-term memory configuration for bridge damage identification using long-term monitoring vibration data. Structural HealthMonitoring, 14759217231161811.

44. Liu, B., Gan, H., Chen, D., & Shu, Z. (2022). Research on Fault Early Warning of Marine Diesel Engine Based on CNN-BiGRU. Journal of Marine Science and Engineering, 77(1), 56.

45. Bao, Y., Tang, Z., Li, H., & Zhang, Y. (2019). Computer vision and deep learning-based data anomaly detection method for structural health monitoring. Structural Health Monitoring, 75(2), 401421.