Accurate deformation prediction is crucial for ensuring the safety and longevity of bridges. However, the complex fluctuations of deformation pose a challenge to achieving this goal. To improve the prediction accuracy, a bridge deformation prediction method based on a bidirectional gated recurrent unit (BiGRU) neural network and error correction is proposed. Firstly, the BiGRU model is employed to predict deformation data, which aims to enhance the modeling capability of the GRU network for time-series data through its bidirectional structure. Then, to extract the valuable information concealed in the error, a transformer model is introduced to rectify the error sequence. Finally, the preliminary and error prediction results are integrated to yield high-precision deformation prediction results. Two deformation datasets collected from an actual bridge health monitoring system are utilized as examples to verify the effectiveness of the proposed method. The results show that the proposed method outperforms the comparison model in terms of prediction accuracy, robustness, and generalization ability, with the predicted deformation results being closer to the actual results. Notably, the error-corrected model exhibits significantly improved evaluation metrics compared to the single model. The research findings herein offer a scientific foundation for bridges’ early safety warning and health monitoring. Additionally, they hold significant relevance for developing time-series prediction models based on deep learning.

Steel beams are widely used load-bearing components in bridge construction. They are prone to internal stress concentration under low-frequency vibrations caused by natural disasters and adverse loads, leading to microcracks and fractures, thereby accelerating the instability of steel components. Therefore, dynamic stress monitoring of steel beams under low-frequency vibrations is crucial to ensure structural safety. This study proposed an external stress sensor based on PMN-PT material. The sensor has the advantages of high sensitivity, comprehensive frequency response, and fast response speed. To verify the accuracy and feasibility of the sensor in actual engineering, the LETRY universal testing machine and drop hammer impact system were used to carry out stress monitoring tests and finite element simulations on scaled I-shaped steel beams with PMN-PT sensors attached. The results show that: (1) The PMN-PT sensor has exceptionally high sensitivity, maintained at 1.716~1.726 V/MPa in the frequency range of 0~1000 Hz. The sensor performance is much higher than that of PVDF sensors with the same adhesive layer thickness. (2) Under low-frequency random vibration, the sensor’s time domain and frequency domain output voltages are always consistent with the waveform of the applied load, which can reflect the changes in the structural stress state in real time. (3) Under the impact of a drop hammer, the sensor signal response delay is only 0.001 s, and the sensitivity linear fitting degree is above 0.9. (4) The simulation and experimental results are highly consistent, confirming the superior performance of the PMN-PT sensor, which can be effectively used for stress monitoring of steel structures in low-frequency vibration environments.

Accurate detection of structural damage is essential for ensuring the safety and reliability of bridges. However, traditional vibration-based approaches often struggle to capture rich feature representations and adequately model spatial dependencies among sensors. This study proposes a novel bridge damage detection framework, LGSTA-GNN, which integrates local–global spatiotemporal learning with graph neural networks. The framework first extracts multi-scale temporal–frequency features using a multi-scale feature extraction module. A local graph feature extraction module then models intrinsic spatial relationships through graph convolutions, while a global graph attention module adaptively captures inter-sensor dependencies by emphasizing structurally informative nodes. A benchmark dataset generated from a scaled bridge model under progressive damage states is used to evaluate the proposed method. Extensive experiments demonstrate that LGSTA-GNN outperforms multiple graph neural network variants and conventional deep learning techniques, achieving superior accuracy, precision, recall, and F1-score. The confusion matrix and t-SNE visualization further verify its enhanced discriminative capability and robustness. Ablation studies confirm the contribution of each module, highlighting the effectiveness of global attention in identifying subtle structural deterioration. Overall, LGSTA-GNN provides an effective and interpretable solution for intelligent bridge damage detection, with strong potential for practical structural health monitoring and real-time safety assessment.

Many cities in coastal areas are prone to typhoon disasters due to their location on the Pacific storm path, and the direct effect of catastrophic winds can lead to the destruction of low-rise buildings and severe damage to high-rise structures. The purpose of this study was to enhance the understanding of boundary layer wind profiles of strong typhoons in coastal areas and reduce property losses and casualties caused by wind disasters. Based on the field measurements of wind profile acoustic radar in coastal areas, the variation characteristics of the boundary layer wind profile during the passage of super typhoon Lekima were first studied in depth, and the evolution law of the typhoon boundary layer profile was summarized. Then, the effects of typhoon horizontal structure, topography, wind speed, and time distance on the characteristics of the typhoon profile were discussed, respectively. Finally, the evolution characteristics of wind profile parameters were obtained by fitting three wind profile theoretical models. Due to the strong variability of typhoon profile morphology, the theoretical model of wind profile is only applicable to the wind profile from the bottom to the low-level jet height of typhoons, while wind parameters are closely related to the spatial location of the typhoon wind field.

Near-fault ground motions significantly threaten bridges due to their distinct features, which are often inadequately considered in current seismic codes based mainly on far-field records. This study analyzes 941 near-fault records to evaluate the effects of site class, pulse-like motions, and vertical components on the peak acceleration ratio and normalized response spectra. A finite element model of a typical simply supported girder bridge is developed to examine how these factors affect pier internal forces. Results show that the peak acceleration ratio increases with softer sites and exhibits large scatter in near-fault regions, indicating that the conventional vertical-to-horizontal ratio of 0.65 may significantly underestimate vertical seismic actions. Pulse motions shift and broaden response spectra, raising seismic demands for medium- to long-period structures. Additionally, pulse effects combined with soft sites cause coupled amplification of internal forces. This work offers a theoretical basis for seismic design and assessment of similar bridges.

Long-term data loss resulting from sensor malfunctions, communication interruptions, and other factors in Structural Health Monitoring (SHM) significantly undermines the reliability of damage identification and safety assessment. Existing methods—ranging from statistical approaches and low-rank matrix completion to traditional machine learning and deep learning imputation techniques—often suffer from either limited accuracy or excessive model size and slow inference, making deployment in resource-constrained scenarios difficult. To address these challenges, this paper proposes TEFN–Imputation, a lightweight and efficient time-series imputation model. This model utilizes observation-driven non-stationary normalization to mitigate the impact of time-varying characteristics and dimensional discrepancies. It employs linear projection for temporal length alignment and constructs BPA-style mass representations from dual perspectives of time and channel. Furthermore, it replaces strict Dempster–Shafer belief combination with an expectation-based evidential aggregation (readout), thereby significantly reducing computational overhead while enabling uncertainty-aware evidential indicators for interpretation rather than claiming a direct accuracy gain from uncertainty modeling. The observed accuracy and robustness improvements are primarily attributed to the normalization and dual temporal–channel modeling design under the same lightweight readout. Systematic experiments on two real-world bridge monitoring datasets, Z24 and Hell Bridge, demonstrate that TEFN consistently maintains low Mean Absolute Error (MAE) and minimal volatility across various combinations of training and testing missing rates, exhibiting high robustness against variations in missing rates and train–test mismatches. Concurrently, compared to RNN and large-scale Transformer baselines, TEFN reduces parameter count and CPU inference time by one to two orders of magnitude. Thus, it achieves a superior trade-off among accuracy, efficiency, and model scale, making it highly suitable for online SHM and imputation tasks in practical engineering applications. Across the settings on Z24, TEFN achieves a mean MAE of 0.218 with a standard deviation of 0.002, while using only 0.02 MB parameters and 2.73 ms per batch CPU inference.