Structural Control & Health Monitoring最新文献

When identifying damage to an offshore wind turbine (OWT) support structure, the influence of harmonic components in vibration response and the difficulty of acquiring data in the damaged state will be encountered. Therefore, the current paper employs the variational mode decomposition (VMD) and sim-to-real deep transfer learning (TL) to identify the damage to an OWT support structure. To eliminate the effect of harmonic components, the vibration response is decomposed using VMD, and the modal response’s reconstructed signal (only containing the structure’s natural frequency) is selected for damage identification. The numerical simulation data and the model test’s measured data are utilized as the source domain (SD) and target domain (TD), respectively. The source model is established by training a convolutional neural network (CNN) with the SD data. The source model’s network structure and weight are frozen to the TD network’s corresponding position. The measured data are utilized to fine-tune the parameters to establish a target model, which is tested to attain the damage identification outcomes. The presented method is validated using the model test data of an OWT support structure.

In this paper, a flutter prediction method based on flutter margins is proposed for streamlined bridges whose aeroelastic behavior can be well approximated by a two-degree-of-freedom (2-DoF) flutter model. The method enables prediction of the flutter boundary by extrapolating a curve of flutter margin versus wind speed, constructed using flutter margins at several subcritical wind speeds. In addition, an $$-optimal flutter active control method based on measured receptances and flutter margins is proposed. It enables assignment of the flutter boundary to a prescribed wind speed value or range for each considered angle of attack (AoA), while simultaneously minimizing vibration responses at subcritical wind speeds, using a single controller with optimal control effort. Hence, the designed controller is robust to the variations of wind speed and AoA. The proposed flutter prediction and control methods require only a small number of systems’ modal parameters or open-loop receptances at several subcritical wind speeds. The proposed flutter prediction method typically requires fewer system modal parameters than existing methods that track the variation of damping ratio against wind speed. The proposed flutter suppression method avoids some modeling errors associated with conventional system matrix-based methods. The working of the proposed flutter prediction and control methods are validated using wind tunnel tests and CFD simulations, respectively.

In tunnel structural health monitoring (SHM) systems, data completeness and accuracy are essential for tasks such as damage detection and early warning. However, environmental disturbances and sensor faults often cause significant missing data, making effective imputation a critical preprocessing step. Traditional statistical methods struggle to capture complex nonlinear temporal and cross-feature dependencies, while autoregressive models, such as recurrent neural networks, suffer from error accumulation and difficulty adapting to dynamically varying strain distributions in real tunnels. To address these challenges, this work proposes a novel nonautoregressive imputation framework based on diffusion models, which effectively mitigate error accumulation. The model effectively exploits the informative content of observed data to guide the modeling and reconstruction of missing values. A gated temporal-feature self-attention fusion module is introduced to accurately capture the complex temporal and spatial dependencies of structural responses. Additionally, external environmental variables such as temperature and water level are integrated to jointly model structural responses and operating conditions, ensuring that the imputation remains robust even under harsh environmental conditions. The method is validated on two real-world SHM datasets from tunnels in Nanjing and Wuhan with various missing data patterns. Experimental results show consistently robust and superior performance across different missing rates, maintaining high accuracy even under severe data loss, demonstrating its effectiveness and practical value in real SHM applications.

Due to their high flexibility and low damping, long-span suspension bridges are susceptible to multimodal vortex-induced vibration (MVIV) under low and normal wind speeds. However, it remains a challenge for the currently used control strategies to achieve optimal additional damping ratios for different modes of vibration as wind speed varies. To this end, this study presents a multimodal control strategy for mitigating MVIV of long-span suspension bridges using magnetorheological (MR) dampers. A vortex-induced force (VIF) model is first established based on the VIFs identified from the wind and structural responses of a bridge during MVIV measured on site. The MVIV of the bridge is then simulated by applying the VIF model to the finite element model of the bridge, and the optimized setup of the control system, consisting of MR dampers and supporting brackets, is sought in terms of a passive control strategy. The multimodal control strategy, which is a novel semiactive control strategy, is finally proposed based on the self-excited characteristics of MVIV observed on site and a linear quadratic regulator. To demonstrate the effectiveness and robustness of the proposed control strategy, a real long-span suspension bridge once suffering MVIV is chosen as a case study. The results demonstrate that the proposed control strategy can robustly mitigate the MVIV of the bridge in the first fourteen modes of vibration in vertical direction, and the effectiveness of the proposed strategy is superior to passive or other semiactive control strategies.

This study investigates the effectiveness of different monitoring strategies for estimating bridge displacement trends induced by landslides, with a focus on addressing three key questions: (i) Can bridge displacement trends induced by a landslide be monitored using only 1D displacement time series along the satellite line of sight (LOS), as provided by InSAR? (ii) How do InSAR-derived displacement trend estimates differ from those obtained through traditional topographic monitoring? (iii) Can a data fusion approach, integrating both InSAR and topographic data, provide more accurate results than using either method alone? Topographic monitoring, which offers direct three-dimensional measurements, is used as the “ground truth” for evaluating the accuracy of InSAR and data fusion methods. The results show that, even though only SAR images from a single orbital geometry are available, InSAR can provide reasonably accurate estimates along the slope-aligned direction, while it is less effective in capturing transverse displacements due to the limitations of measuring along the satellite’s LOS. However, when combined with prior knowledge of landslide behavior, InSAR still provides valuable insights. Bayesian data fusion, which integrates topographic and InSAR measurements, significantly reduces uncertainties, particularly in short monitoring periods, offering a cost-effective alternative to continuous topographic monitoring. Additionally, this study explores two alternative strategies: limiting topographic measurements to the first year and spreading sparse topographic measurements over several years and relying on satellite data thereafter. While both approaches yield satisfactory results in the slope direction, they show higher uncertainties in the transvers direction, particularly as the frequency of topographic measurements decreases. The findings suggest that a combined monitoring approach, integrating satellite and topographic data, as well as a prior knowledge of landslide behavior, provides an accurate and cost-effective solution for long-term monitoring of infrastructure in landslide-prone areas.

A novel data-driven damage detection framework, based on probabilistic modeling of mel-frequency cepstral coefficient (MFCC) distributions, is proposed. This study replaces traditional autospectral densities with the power spectrum derived from cross-spectral density ratios in MFCC extraction, enhancing sensitivity to subtle dynamic changes across structural locations. The Bhattacharyya distance (D_B) is introduced to quantify dissimilarities between MFCC distributions under baseline and potential damage scenarios. Subsequently, a damage indicator (DI_B) based on the Bhattacharyya distance is proposed. To mitigate uncertainties caused by environmental noise and measurement variability, a statistically sound threshold is established through Bayesian resampling and Monte Carlo simulation. When the DI_B values of structural states exceed this threshold, it indicates the presence of damage. Additionally, a vectorization scheme is employed to improve computational efficiency, enabling faster processing of multichannel data. The accuracy and effectiveness of the proposed method are validated through a laboratory experiment involving four beams and a field test conducted on a steel truss bridge. The results demonstrate the proposed method’s ability to detect and classify damage states accurately under diverse conditions, highlighting its applicability for reliable structural health monitoring (SHM).

Developing fully automated operational modal analysis (AOMA) algorithms is a critical yet challenging task in structural health monitoring, with urgent practical engineering demands. This study introduces a robust AOMA framework that leverages an enhanced stabilization diagram and hierarchical density estimation strategy to address these challenges. The main innovations of this framework are threefold: (1) A comprehensive physical mode validation strategy that effectively eliminates more stubborn spurious poles by destroying the noise structure inherent in the identified system matrix. (2) A hierarchical density clustering approach for the automatic interpretation of stabilization diagrams, which eliminates the need for manual threshold selection and adapts seamlessly to varying-density clustering scenarios. (3) A novel representative mode selection approach based on clustering exemplars is presented, resulting in a stronger consistency of the selected modal parameters. Hierarchical clustering of modal poles, optimal cutting of clustering trees, outlier rejection, and cluster quality validation are integrated in a single framework, streamlining the analysis and avoiding tedious postprocessing steps. The robustness and applicability of the algorithm are extensively validated using a numerical building structure, the Z24 bridge benchmark test, and a footbridge equipped with a long-term continuous monitoring system. The results demonstrate that the proposed framework achieves robust AOMA on long-term field measurement data without any user intervention. The applicability of the algorithm to closely spaced modes and long-term modal tracking tasks is also demonstrated. This study advances the field of AOMA by offering a scalable, efficient, and accurate solution for real-time structural health assessment, with potential extensions to broader engineering applications.

In this article, two advanced imaging-based metrology methods, the multipoint method and the tele-wide-angle method, are introduced to the field of structural health monitoring. Both provide the means to significantly improve either the measurement uncertainty or the field of view compared to classical imaging-based methods. The multipoint method utilizes a computer-generated hologram to replicate a single object point to a predefined spot pattern in the image. Spatial averaging of the spot positions improves the measurement uncertainty. The second method, called tele-wide-angle, uses a diffraction grating to considerably enlarge the field of view of a tele objective lens. Both methods are investigated regarding the achievable measurement uncertainty at distances between 34 and 50 m. The standard deviations of the error range between 0.027 and 0.034 mm for the multipoint method and 0.008 and 0.02 mm for the tele-wide-angle method. In the second part of the article, both measurement systems are employed in a field study, measuring the deformation of a railway bar arch bridge. An inductive displacement transducer and several accelerometers are installed to validate the measured displacements and dynamics.

To overcome the challenge where convolutional neural networks (CNNs) struggle to effectively capture the diverse visual features of cracks, spalling, and exposed rebar in concrete structures resulting in inaccurate damage segmentation, a method is proposed, known as the cross-domain coupled convolutional transformer network for concrete damage detection (DamageNet). First, a dual-branch encoder architecture combining CNN and transformer is designed with a hierarchical structure that outputs CNN and transformer features at the same resolution, preserving both local perception and global information. Second, a cross-domain coupling attention module is introduced to integrate the CNN and transformer features effectively, fusing local perception and global modeling information in a complementary manner. Finally, on multiple publicly available multidamage datasets, the proposed network achieves IoU scores of 78.70%, 91.52%, and 73.90% for exposed rebar, cracks, and spalling, respectively, and the mean ± standard deviation across all damage classes obtained from five training repetitions is 82.74% ± 2.46%. Experimental results validate that the proposed network outperforms other mainstream methods, and the feature map visualization demonstrates that the network effectively captures diverse visual features, benefiting concrete multidamage detection tasks.

Structural health monitoring (SHM) data are crucial for structural state assessment. However, long-term monitoring data are inevitably subject to data missing in actual SHM, which seriously hinders the reliability of the SHM system. So far, many deep learning-based supervised data imputation methods have been proposed, which require complete sensor data for training. Although there are studies on unsupervised data imputation, some complete sensor data are still required. Especially, there is a lack of study on the challenging problem of unsupervised data imputation with incomplete data of all sensors, which may occur in actual SHM. Therefore, an enhanced generative adversarial imputation network with unsupervised learning is proposed in this paper for such a challenging task. First, within the generative adversarial imputation network framework, convolutional neural networks (CNNs) with an encoder–decoder architecture are established to extract significant high-level local features. Furthermore, a self-attention mechanism is embedded into the generative network to globally capture remote dependencies between data. Finally, the skip connections are incorporated to enhance the parameter utilization and imputation performance of the network. The random missing data imputation with incomplete data of the field monitoring acceleration data from the Dowling Hall footbridge is used to validate the proposed method. The results show that good data imputation in both the time and frequency domains can be achieved by the proposed method in the case of random data missing in all sensors.