Structural Control & Health Monitoring最新文献

{"title":"A Geometry-Aware Crack Width Quantification Model for Concrete Structures Based on Attention-ResNet-UNet Segmentation","authors":"Jiehui Wang,&nbsp;Ziyang Chen,&nbsp;Hongyuan Guo,&nbsp;Jian-Guo Dai","doi":"10.1155/stc/5802814","DOIUrl":"https://doi.org/10.1155/stc/5802814","url":null,"abstract":"<p>Accurate crack width quantification is a fundamental prerequisite for structural safety evaluation, durability assessment, and maintenance decision-making in concrete infrastructures. However, most deep learning–based approaches primarily focus on segmentation accuracy, while their effectiveness in downstream geometric measurement remains limited. This study proposes a geometry-aware crack width quantification framework that links segmentation fidelity to measurement accuracy. This framework integrates an attention-ResNet-UNet segmentation model with a skeleton-guided maximum inscribed circle algorithm. An edge-guided CBAM attention mechanism is introduced to enhance feature representation of fine-scale crack boundaries and suppress background interference, thereby improving segmentation continuity and edge precision. Building upon high-quality segmentation outputs, a geometry-driven width extraction strategy is employed to generate spatially continuous and orientation-invariant crack width profiles along the medial axis. A comprehensive evaluation was conducted on 85 annotated crack images out of a total of 1011 images that were reserved as an independent testing dataset, and the proposed framework was compared with two baseline models, FCN and standard UNet, under identical training conditions. Experimental results demonstrate that the proposed method achieves significantly higher accuracy in width estimation, reducing MAE and RMSE by over 40% compared with the baselines and achieving an <i>R</i>² of 0.93, markedly outperforming FCN (−1.18) and UNet (−3.71). Scatter distribution analysis further shows tighter clustering around the ideal fitting line, indicating improved stability and robustness. The results highlight the necessity of evaluating vision-based structural inspection methods using engineering-oriented metrics rather than segmentation indicators alone. The proposed framework provides an effective approach for preserving crack geometry and improving pixel-level width estimation, while further physical calibration and field validation are required before its application to absolute crack width measurement in real-world infrastructure.</p>","PeriodicalId":49471,"journal":{"name":"Structural Control & Health Monitoring","volume":"2026 1","pages":""},"PeriodicalIF":5.0,"publicationDate":"2026-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/stc/5802814","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148994353","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"ReplightNet: A Novel Lightweight Network Model for Fault Diagnosis of Rotating Machinery","authors":"Yan Li,&nbsp;Jincheng Li,&nbsp;Lei Zhang,&nbsp;Shaohu Liu,&nbsp;Xiangdong Xie,&nbsp;Kai Guo","doi":"10.1155/stc/8261028","DOIUrl":"https://doi.org/10.1155/stc/8261028","url":null,"abstract":"<p>Currently, the majority of existing methods for rotating machinery fault diagnosis are based on single-axis vibration data. It is often difficult to obtain satisfactory diagnostic results under complex operating conditions. Besides, these methods are limited by excessively large model parameters and high computational costs. Such drawbacks make them unable to meet the requirements of practical applications. To overcome these challenges and achieve fast and accurate rotating machinery fault diagnosis, a simple and effective multiaxes information fusion method and a lightweight neural network (ReplightNet) are proposed in this paper. First, the <i>XYZ</i> three-axes vibration signals are fused into color time–frequency maps. Then, ReplightNet is designed for fault diagnosis. Specifically, the basic block is constructed through structural reparameterization and an efficient spatial and channel attention (ESCA) module, combining convolution and pooling to design the efficient downsampling module PCDown, and using Kolmogorov–Arnold networks (KANs) as the classification layer to achieve the fitting of complex network structures with fewer parameters. Finally, extensive experiments are conducted under multiple operating conditions on two types of datasets to validate the effectiveness and adaptability of the proposed method. The test results show that the proposed method is superior in terms of fault diagnosis accuracy and model parameter complexity compared to other advanced diagnostic models.</p>","PeriodicalId":49471,"journal":{"name":"Structural Control & Health Monitoring","volume":"2026 1","pages":""},"PeriodicalIF":5.0,"publicationDate":"2026-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/stc/8261028","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148994394","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Diameter Inversion for Dense Rebars Based on Dual-Polarized GPR","authors":"Shuangying Sun,&nbsp;Junbo Gong,&nbsp;Yuhan Li,&nbsp;Kunao Li,&nbsp;Tian Lan","doi":"10.1155/stc/5470113","DOIUrl":"https://doi.org/10.1155/stc/5470113","url":null,"abstract":"<p>Reinforced concrete is an essential material in building engineering. Estimating rebar diameter is crucial for evaluating structural load-bearing capacity and overall stability, which is of great significance for the structural health monitoring of buildings. Ground-penetrating radar (GPR), as a nondestructive testing technique, can be used to assess rebar diameter. However, in practical engineering, rebars in concrete are often densely distributed, leading to overlapping electromagnetic echoes from multiple targets. Existing diameter inversion methods based on GPR are primarily designed for isolated single rebar and are not well-suited for diameter estimation in multitarget scenarios. To tackle this problem, this paper presents a fast multiecho separation method to improve the accuracy of diameter inversion in multirebar scenarios. First, a multirebar scattering model for rebars is established. Subsequently, through improved deconvolution and the orthogonal matching pursuit algorithm, individual rebar echoes within the overlapping echoes are rapidly resolved. Finally, simulation and practical experiments have verified that this method can effectively separate individual rebar echoes from multiple rebar echoes. By integrating the diameter inversion algorithms based on the wideband power ratio and natural resonance frequency, respectively, the accuracy of diameter inversion can be significantly enhanced. The relative error in the inverted diameter has decreased by approximately 23%–58%, and the stability of the diameter estimation algorithm has been effectively improved. Compared to the echo separation method using original deconvolution and orthogonal matching pursuit, the enhanced multiecho separation method reduces the computation time by about 80%.</p>","PeriodicalId":49471,"journal":{"name":"Structural Control & Health Monitoring","volume":"2026 1","pages":""},"PeriodicalIF":5.0,"publicationDate":"2026-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/stc/5470113","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148994546","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Novel Condition-Monitoring System Based on a Parametric Correlation Learning Network for Wind Turbine Blade Fault Identification","authors":"Vijay Tiwari,&nbsp;Lalit Kumar Bansal","doi":"10.1155/stc/1499553","DOIUrl":"https://doi.org/10.1155/stc/1499553","url":null,"abstract":"<p>The role of wind turbine blades in determining the reliability and performance of wind energy systems is critical. However, environmental and operational stressors, as well as wind shear, can damage blades. The safe and effective operation, free of failures and with reduced blade maintenance costs, requires the reliable assessment and detection of blade damage. The existing damage detection methodologies, however, fail to recognize operational turbine acoustic emissions due to their complexity, uncharacteristic and nonlinear behaviour, particularly under noisy, varying operational conditions. An attention-enhanced temporal convolutional network (AMTCN) and a bidirectional gated recurrent unit (AMTCN-BiGRU) are proposed in this study for analysing acoustic time-series data. Unlike conventional hybrid TCN-RNN architectures, the proposed framework is designed to hierarchically learn fault-sensitive acoustic correlations by combining attention-guided multiscale temporal feature extraction with bidirectional sequential dependency modelling for improved blade fault discrimination. Improving feature representation and more effectively capturing temporal dependencies is the objective of the integrated attention, temporal convolution and bidirectional sequence processing. This formulation fits the damage identification scenario with limited, noisy datasets, where acoustic signals exhibit substantial variation. The proposed methodology has been benchmarked against other models such as TCN, TCN-LSTM and AMTCN-GRU. The results showed that the AMTCN-BiGRU model outperformed the other models, achieving an RMSE of 0.1369, an MAE of 0.059 and an <i>R</i>² of 0.990 with 97.3% accuracy. The findings show the effectiveness of the proposed method for detecting damage at early stages and for structural health monitoring of wind turbine blades.</p>","PeriodicalId":49471,"journal":{"name":"Structural Control & Health Monitoring","volume":"2026 1","pages":""},"PeriodicalIF":5.0,"publicationDate":"2026-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/stc/1499553","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148994550","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"H2 Optimization of Tuned Liquid Column Damper With Inerter and Negative Stiffness According to Different Seismic Response Objectives","authors":"Jue Wang,&nbsp;Nuo Xu,&nbsp;Yongxin Wu,&nbsp;Daniel T.W. Looi","doi":"10.1155/stc/7132597","DOIUrl":"https://doi.org/10.1155/stc/7132597","url":null,"abstract":"<p>A nonlinear tuned liquid column damper with inerter and negative stiffness (NS-TLCDI) is proposed for structural vibration control with frequency tunability. The close-form transfer functions of different responses are presented by introducing an equivalent linearization method. The Routh–Hurwitz stability criterion is applied to confirm the stability of system with negative stiffness. Subsequently, analytical solutions for the <i>H</i>₂ optimized frequency ratio and damping ratio of NS-TLCDI under a white noise excitation are derived based on different engineering requirements that prioritize either absolute acceleration or relative displacement, the validity of which is verified by comparison with corresponding numerical solutions. Furthermore, the vibration suppression of NS-TLCDI in the frequency domain is assessed compared with a tuned liquid column damper inerter (TLCDI). The results indicate that NS-TLCDI offers a significant advantage over TLCDI in reducing the mean square values of the dynamic responses of the main structure, which becomes more pronounced as the absolute value of negative stiffness increases within the stability requirements. Moreover, the robustness of the system equipped with NS-TLCDI has also been considerably enhanced under this trend. Finally, a detailed time-history response analysis is conducted to evaluate the vibration control performance of NS-TLCDI under 44 real earthquake excitations. Both in terms of peak values and overall, the time-history responses of the structure equipped with optimal NS-TLCDI are consistently lower than that of optimal TLCDI. The energy dissipation of systems is further examined, indicating that the dashpot of NS-TLCDI remains the dominant dissipation component.</p>","PeriodicalId":49471,"journal":{"name":"Structural Control & Health Monitoring","volume":"2026 1","pages":""},"PeriodicalIF":5.0,"publicationDate":"2026-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/stc/7132597","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148994241","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Interpretable Spatiotemporal Forecasting and Probabilistic Working-Condition Assessment of Prestressed Anchor Cables in Reinforced Wind-Turbine Foundations","authors":"Jizhang Zhao,&nbsp;Huanwei Wei,&nbsp;Xiangrong Gao,&nbsp;Fang Tan,&nbsp;Xingkuo Wang,&nbsp;Xiao Zheng","doi":"10.1155/stc/4897404","DOIUrl":"https://doi.org/10.1155/stc/4897404","url":null,"abstract":"<p>As the service life of onshore wind turbines increased, their concrete foundations experienced progressive deterioration, making post-reinforcement condition assessment increasingly important. However, studies on the future response of prestressed anchor cables in reinforced wind-turbine foundations remained limited. To address this issue, an interpretable data-driven framework was developed for anchor-cable axial-force forecasting and probabilistic working-condition assessment. First, XGBoost combined with SHAP analysis was used to identify the dominant variables associated with anchor-cable axial-force prediction and to construct reduced feature sets. Compared with models trained using the full feature set, the reduced feature sets achieved an average training speedup of 1.86-fold and an average relative improvement of 35.5% in <i>R</i>². Subsequently, an STA-Transformer was developed by integrating temporal dependency modeling, multiscale feature extraction, adaptive feature enhancement, and global self-attention. Ablation experiments verified the contributions of the incorporated modules. Compared with iTransformer, TimesNet, Autoformer, and Informer, the STA-Transformer achieved average <i>R</i>², mean squared error, root mean squared error, and mean absolute error values of 0.970, 0.211, 0.431, and 0.327, respectively. Finally, multistep forecasting was coupled with Monte Carlo simulation to evaluate the future working conditions of the prestressed anchor cables. The predicted axial-force distributions showed reasonable agreement with the field measurements over the evaluated prediction horizons. The proposed framework thereby provided an effective tool for anchor-cable working-condition assessment and could support the operation and maintenance of reinforced wind-turbine foundations.</p>","PeriodicalId":49471,"journal":{"name":"Structural Control & Health Monitoring","volume":"2026 1","pages":""},"PeriodicalIF":5.0,"publicationDate":"2026-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/stc/4897404","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148994242","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Novel Ice-Plug Temporary Sealing Structural System for Deep Outlet Facilities: From Engineering Concept to Feasibility Assessment","authors":"Cao Xiangyu,&nbsp;Tang Yunqing,&nbsp;Qian Long,&nbsp;Liu Haixiang,&nbsp;Hu Songdi,&nbsp;Chen Xining,&nbsp;Gao Panfeng,&nbsp;Ke Minyong","doi":"10.1155/stc/5778852","DOIUrl":"https://doi.org/10.1155/stc/5778852","url":null,"abstract":"<p>High-head deep outlets require temporary sealing methods that are reversible, rapidly deployable and mechanically reliable. This study proposes a novel in situ ice-plug structural system in which tunnel water upstream of the gate is frozen to form a temporary barrier, while the ice plug, concrete lining and surrounding rock act as a confined composite system to resist hydrostatic loading. The system is characterised as a load-transfer and stability problem governed jointly by ice compression, interfacial shear, lateral confinement and plug geometry, rather than by ice strength alone. The feasibility of the concept is evaluated through an analytical equilibrium formulation, two-stage physical model tests and complementary three-dimensional heat-transfer and nonlinear mechanical simulations. The tests demonstrate controllable condenser-pipe ice formation and provide pressure–displacement data for equivalent-parameter calibration, while identifying end sealing and ice–lining contact as critical constraints. Heat-transfer analysis shows that convection at the ice–water interface dominates cold-energy loss, whereas conduction into the lining and surrounding rock is secondary. Mechanical analyses reveal a predominantly compressive stress state and show that antisliding capacity is controlled primarily by interfacial resistance, surrounding-rock confinement and the plug length-to-diameter ratio. Under the adopted conditions, <i>L</i>/<i>D</i> = 0.4 is suggested as a preliminary lower bound for heads not exceeding 100 m. These findings clarify the governing thermal and mechanical mechanisms of the ice-plug–lining–rock system and provide a theoretical basis for preliminary design and further field-scale assessment.</p>","PeriodicalId":49471,"journal":{"name":"Structural Control & Health Monitoring","volume":"2026 1","pages":""},"PeriodicalIF":5.0,"publicationDate":"2026-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/stc/5778852","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148994477","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"VisionCrack: Smartphone-Based Fine Crack Width Measurement via Adaptive Scaling With Multithickness 1D Barcodes","authors":"Ruslan Raupov,&nbsp;Hyun-Ki Kim,&nbsp;Sung-Jin Lee,&nbsp;Hojune Ann","doi":"10.1155/stc/7069645","DOIUrl":"https://doi.org/10.1155/stc/7069645","url":null,"abstract":"<p>Accurate quantification of surface crack width is essential for maintaining civil infrastructure, yet conventional visual inspection is subjective and inconsistent. This study presents VisionCrack, a smartphone-based system for low-cost image-based crack width measurement using a 1D barcode. By analysing five predefined barcode bar thicknesses (T1–T5), VisionCrack estimates an adaptive scale (mm/pixel) using linear regression. Rather than relying on a single fixed reference, the method uses multiple barcode thickness references within each image to compensate for measurement bias. In experiments against crack width gauges, VisionCrack achieved a mean absolute error (MAE) of 0.0105 mm. Compared with a single-reference fixed-scale baseline representing prior marker-based approaches, the proposed method showed a lower error (0.0105 mm vs 0.1183 mm). These results demonstrate the calibration accuracy and practical potential of the proposed method under controlled image-capture conditions.</p>","PeriodicalId":49471,"journal":{"name":"Structural Control & Health Monitoring","volume":"2026 1","pages":""},"PeriodicalIF":5.0,"publicationDate":"2026-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/stc/7069645","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148994475","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Interpretable Machine Learning Stacking Ensembles for Dam Displacement Prediction and Safety Assessment","authors":"Zeghmar Amer,&nbsp;Chebbah Lynda,&nbsp;Kabour Abdesselem,&nbsp;Belkendil Abdeldjalil,&nbsp;Yahi Takai Eddine,&nbsp;Aqil Tariq,&nbsp;Muhammad Sohail Waqas,&nbsp;Dmitry Kucher,&nbsp;Vladimir Dmitriev,&nbsp;Saad S. Alarifi","doi":"10.1155/stc/6611597","DOIUrl":"https://doi.org/10.1155/stc/6611597","url":null,"abstract":"<p>Monitoring dam movements is essential to ensure structural safety and to plan maintenance interventions. However, dam displacement is influenced by multiple interacting factors, including reservoir water-level fluctuations, air–water–concrete temperatures, wind speed, and relative humidity, making its behavior highly complex and nonlinear. To address this challenge and improve short-term displacement forecasting for the Beni Haroun Dam in Algeria, this study develops a machine learning–based stacking ensemble model. Five regression algorithms, Random Forest (RF), Extra Trees (ET), Extreme Gradient Boosting (XGBoost), Categorical Boosting (CatBoost), and Multiple Linear Regression (MLR), were first trained using a fivefold cross-validation procedure. Their out-of-fold predictions were then combined as metafeatures to build a higher-level stacking ensemble model, with CatBoost employed as the metalearner. Model performance was evaluated using NSE, RMSE, <i>R</i>², MAE, PBIAS, and MAPE. The results demonstrate the clear superiority of the stacking approach, which achieved <i>R</i>² = 0.971, RMSE = 0.292, MAE = 0.236, NSE = 0.971, and near-zero bias (PBIAS = 0.000). This represents a substantial improvement over the best-performing individual model, ET (<i>R</i>² = 0.843, RMSE = 0.686), as well as over other models such as CatBoost (<i>R</i>² = 0.831), RF (<i>R</i>² = 0.811), MLR (<i>R</i>² = 0.803), and XGBoost (<i>R</i>² = 0.797). The MLR model showed comparatively lower performance (<i>R</i>² = 0.803, RMSE = 0.768), confirming the advantage of nonlinear approaches. SHAP analysis identified reservoir water level as the most influential factor in predicting dam displacement. The results confirm that the stacking ensemble model effectively captures complex dam behavior, providing accurate forecasts that support early anomaly detection and improved dam safety management.</p>","PeriodicalId":49471,"journal":{"name":"Structural Control & Health Monitoring","volume":"2026 1","pages":""},"PeriodicalIF":5.0,"publicationDate":"2026-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/stc/6611597","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148994476","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Self-Attention-Enhanced TCN-BiGRU Framework for Missing Temperature Data Recovery in Construction-Stage Bridge Monitoring","authors":"Zujun Zhang,&nbsp;Jiangyan Wu,&nbsp;Guoliang Zeng,&nbsp;Wenping Peng,&nbsp;Zhongchu Tian,&nbsp;Yang Deng,&nbsp;Benkun Tan","doi":"10.1155/stc/7578661","DOIUrl":"https://doi.org/10.1155/stc/7578661","url":null,"abstract":"<p>Accurate monitoring data during bridge construction are critical for structural safety analysis. However, data loss and distortion caused by sensor or power faults occur frequently. This study proposes an integrated model combining temporal convolutional networks (TCNs), bidirectional gated recurrent units (BiGRUs), and a self-attention (SA) mechanism (TCN-BiGRU-SA) to address this issue by formulating missing-data recovery as a reconstruction task. To this end, raw signals are denoised via wavelet thresholding, and training samples are constructed using sliding windows. Subsequently, missing-data patterns and proportions are simulated via masking to train the model. A comparative validation analysis using construction-stage monitoring data from an arch bridge and considering standalone TCN, BiGRU, and hybrid models shows that, at a 50% missing rate, the proposed model improved <i>R</i>² by approximately 0.25%–20.01% compared with the six benchmark models and decreased the root-mean-square error by 12.5% for 30% continuous missing data at the east-oriented measurement point of the No. 22 tensioning cable, providing an efficient and accurate approach for monitoring-data recovery.</p>","PeriodicalId":49471,"journal":{"name":"Structural Control & Health Monitoring","volume":"2026 1","pages":""},"PeriodicalIF":5.0,"publicationDate":"2026-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://onlinelibrary.wiley.com/doi/epdf/10.1155/stc/7578661","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148994140","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}