Structural Control & Health Monitoring最新文献

A novel structural damage detection (SDD) method is proposed in this work, which is based on adversarial-based transfer learning to achieve the cross-domain information transfer of damage locations between numerical simulations and real bridge structures. Although the advancement of numerical modeling technology makes it accessible for relatively accurate finite element (FE) models, it is still difficult to meet the needs of practical engineering. The idea of adversarial training is introduced to enable the traditional feature extraction network to obtain the domain independent features between the numerical simulation and the real bridge structure. The dynamic response data from the numerical simulations are labeled with damage, while those from the real structure are unlabeled. To verify the effectiveness of the proposed method, we established a FE model of a simply support beam and regarded it as the benchmark model, and the target model with discrepancies from the benchmark model is obtained by quantitatively increasing the uncertainties. The results of the simulation show that the proposed method can overcome the discrepancy caused by uncertainty to a certain extent compared with the traditional method and obtain a high damage localization accuracy on the target model. In the laboratory experiment, the proposed method still achieves promising results. The primary contributions of this work are twofold: first, it delves deeper into the effectiveness of adversarial training for extracting domain-invariant features, which are crucial for structural damage identification. Second, it provides a quantitative assessment of the performance degradation of traditional methods due to modeling errors and uncertainty. Additionally, it demonstrates the significant performance enhancement achieved by the proposed method.

This paper presents an on-site anchor bolt tension-force versus deformation detection method, which enables the assessment and evaluation of the extent of damage in the anchored zone of anchor bolts buried deep within concrete. In addition, considering different toppling modes and their respective force mechanisms, the paper proposes using the entire cross-sectional plastic bending moment with a 0.4 impact coefficient as the actual effective bending moment for the anchor bolt foundation. This is employed to calculate the tensile forces in the anchor bolts and the local compressive bearing capacity of the concrete above the lower anchor plate. The computational results indicate that the plastic damage in the anchored zone of problematic anchor bolts is primarily caused by the crushing of concrete above the lower anchor plate rather than axial plastic deformation of the anchor bolts. Accordingly, targeted reinforcement schemes for repairing lower anchor plate deformations and replacing locally crushed concrete are proposed and successfully applied, providing practical reference for engineering applications.

The sensor-based dam health monitoring (DHM) systems of concrete-faced rockfill dam (CFRD) are easily affected by environmental factors, which inevitably causes sensor fault, and the measured value of its effect quantities is nonlinear and unstable. The application of machine learning in the preprocessing of dam safety monitoring data is very extensive, mainly including two parts: gross error elimination and missing data completion. In this paper, support vector regression (SVR), a typical machine learning algorithm, is chosen to accomplish these two tasks, while suggesting possible optimizations in different situations of hydraulic monitoring, including optimization of parameters in SVR using the population algorithm sparrow search algorithm (SSA); optimization of the pattern of gross error discriminant using the minimum covariance determinant (MCD) algorithm; and the hierarchical clustering on principal components (HCPC) algorithm to optimize the selection method of spatial measurement points when completing a segment of missing data. The results show that the optimized SVR method has greater accuracy in both gross error elimination and the completion of individual missing data or a segment of missing data for DHM systems, which is applicable to measured data of CFRD. These optimization methods can also be extended to other engineering applications.

Digital twins (DTs) have revolutionised digitalisation practices across various domains, including the Architecture, Engineering, Construction and Operations (AECO) sector. However, DTs often face challenges related to data scarcity, especially in AECO, where tests are costly and difficult to scale. Historical data in this domain are often limited, unstructured and lack interoperability standards. Data scarcity directly affects the accuracy and reliability of the DT models and their decision-making capabilities. To address these challenges, classical methods are used to produce synthetic data based on predefined statistical distributions, which are barely scalable to unpredictable scenarios and prone to overfitting. Alternately, this work presents a novel comprehensive approach that covers every aspect from synthetic data generation to training and testing of these data on the system’s models. This strategy not only delivers high-quality data that meets the model’s requirements in terms of diversity, complexity and class balance, but also provides the diagnostic and prognostic capabilities of the DT of the system through its trained models. State-of-the-art techniques including generative adversarial networks (GANs), specifically Wasserstein generative adversarial networks with gradient penalty (WGAN-GP), and convolutional neural networks (CNNs) are employed in this novel pervasive approach, participating in the same architecture for generative, diagnostic and prognostic purposes. GANs enable data augmentation and reconstruction, while CNNs excel in spatial pattern recognition tasks. The proposed framework is demonstrated through an experimental case study on damage diagnostics and prognostics of a laboratory-scale metallic tower, where synthetic datasets are generated to supplement limited health monitoring data. The results showcase the effectiveness of the generated data for damage detection, prognostics and operational decision-making within the DT context. The presented method contributes to overcoming data scarcity challenges and improving the accuracy of DT models in the AECO sector. The article concludes with discussions on the application of the results and their implications for decision-making within the DT framework.

Accurate estimation of modal parameters is crucial for various aspects of tall buildings, including structural design, vibration control, and state assessment. This paper first presents a combined approach for the structural modal parameter estimation by combining the empirical wavelet transform (EWT), smoothed discrete energy separation algorithm-1 (SDESA-1), and half-cycle energy operator (HCEO), referred to as EWT-SH. A numerical study on a five-story frame structure is conducted using the Newmark-β method to validate its effectiveness and accuracy. The results demonstrate that relative errors in estimating the natural frequency and damping ratio using the EWT-SH method are significantly smaller compared to traditional methods. Furthermore, the EWT-SH method is applied to estimate the modal parameters of a real super-tall building, i.e., the SEG Plaza building in Shenzhen, using acceleration responses. Identified results confirm the applicability and accuracy of the EWT-SH method in real-world scenarios and indicate that the frequencies and damping ratios of the SEG Plaza building noticeably decrease after 20 years of service, which could partially explain the SEG building vibration event on May 18, 2021. Since the identified frequencies are quite different from those of the original finite element (FE) model of the tall building, the dual-loop particle swarm optimization (PSO) is specifically developed to update the FE model of SEG Plaza building.

Tuned mass damper (TMD) is a single-terminal damper, while inerter is a two-terminal one, which are effective control devices. So, a hybrid damper with a series-parallel inerter system and a TMD (SPIS-TMD) in series is proposed. The main work of the manuscript is as follows. Firstly, based on the mechanical diagram of SPIS-TMD and its equipment on the roof of a high-rise structure, the general form of seismic motion equation was derived using dynamic finite element technology. Secondly, based on the method of quadratic decomposition for power spectrum density function (QD-PSDF), the concise analytic solutions for zero-, first-, and second-order response spectral moments (ZFSO-RSMs) of the general response of structure with SPIS-TMD were deduced. Thirdly, in response to the numerous parameters that affect the safety of high-rise structures and the varying difficulty for obtaining SPIS-TMD’s parameters, an optimization analysis technique was proposed, which is constrained by dynamic reliability and SPIS-TMD’s parameter weights. Finally, three examples were given and results show the following. (1) The proposed analytical solutions for ZFSO-RSMs are correct and high efficiency and can be extended to analysis of stationary seismic responses of general linear structures. (2) Calculating results of ZFSO-RSMs of general responses with the number of vibration modes corresponding to a participation weight of 100% is exactly the same as those with all vibration modes, and the calculating time of the case of the participation weight of 100% is less than 1/12 that of the case of all vibration modes. (3) The failure probabilities of the structure with SPIS-TMD with optimal parameters using the proposed method (with a limit failure probability of 0.1587), the structure only equipped with TMD with the same TMD’s parameters as SPIS-TMD, and the structure without dampers are 0.1570, 0.7060, and 0.9778, respectively. It indicates that under the same conditions, the hybrid damper SPIS-TMD has a better damping effect than a single TMD.

A smart concrete girder bridge usually has various sensors, based on which several physical properties can be measured, and hence, the health condition can be evaluated. Cracks are always observed on a smart concrete girder bridge. In particular, some of the cracks are induced by overloaded vehicles, which is dangerous to its safe operation. However, due to the crack opening and closing effect, it exhibits nonlinear responses, posing challenges for accurately assessing its health condition. Influence lines (ILs) are a promising indicator for bridge damage. However, there is limited research on the effect of cracks on the ILs of a smart concrete girder bridge. A digital twin is commonly used to accompany the smart sensing system to accurately evaluate the health condition, where the finite element (FE) model is of great importance. Therefore, this study proposes an element size–independent FE model construction method based on the concrete damage plasticity (CDP) model to investigate the changes of displacement and strain ILs of different types of smart concrete girder bridges with bending cracks, which is helpful to guide how to use the ILs to identify the cracks and evaluate the health condition. Initially, a concrete constitutive model based on crushing/fracture energy is proposed, and the evolution law of tensile damage based on fracture energy is derived to construct the element size–independent FE model. Subsequently, experiments on a reinforced concrete (RC) simply supported beam and a prestressed concrete (PC) simply supported bridge subjected to bending failure are used to verify the FE models constructed by the proposed method. Finally, the FE models of a smart RC T-beam bridge and a smart three-span PC continuous bridge are established to study the changes in ILs caused by bending cracks. The change of displacement IL at the midspan due to cracks for the smart RC bridge exceeds 10% when the reinforcements yield, while it is less than 10% for the smart PC bridge even if the bridge is in the failure state. The change of both displacement and strain ILs becomes greater when the measurement point approaches the cracks, and the change of strain IL is only detectable when the measurement is close to the cracks. Due to the crack opening and closing effect, the displacement and strain ILs of a smart concrete girder bridge with bending cracks are inconsistent when different loads are applied. The findings can also be used as a pre-IL-based crack detection using the passing inspection vehicle-induced dynamic response on a selection of type of ILs, determination of layout of sensors, and mass of inspection vehicle.

Deformation is a critical indicator for the safety control of high-arch dams, yet traditional statistical regression methods often exhibit poor predictive performance when applied to long-sequence time series data. In this study, we develop a robust predictive model for deformation behavior in high-arch dams by integrating signal dimensionality reduction with deep learning (DL)-based residual correction techniques. First, the fast Fourier transform is employed to decompose air and water temperature sequences, enabling the extraction of temperature cycle characteristics at the dam boundary. A data-driven statistical monitoring model for dam deformation, based on actual temperature data, is then proposed. Subsequently, an improved Bayesian Ridge regression model is used to construct the dam deformation monitoring framework. The residuals that traditional statistical methods fail to capture are input into an enhanced Long Short-Term Memory (LSTM) network to effectively learn the temporal characteristics of the sequence. A high-arch dam with a history of long-term service is used as a case study. Experimental results indicate that the data dimensionality reduction method effectively extracts relevant information from observed temperature data, reducing the number of input variables. Comparative evaluation experiments show that the proposed hybrid predictive model outperforms existing state-of-the-art benchmark algorithms in terms of predictive efficiency and accuracy. Additionally, this approach combines the interpretability of statistical regression methods with the powerful nonlinear modeling capabilities of DL-based models, achieving a synergistic effect.

The bearing capacity of the existing fine-grained soil subgrade is mainly achieved by measuring the modulus, and its testing methods can be divided into two categories including static method and dynamic method. The data connection between the two is still lacking in systematic research. The traditional BB (Benkelman Beam) static test has disadvantages such as slow detection speed, low accuracy, and poor reliability due to the use of deflection index, which is not fully applicable to the expressway. To this end, dynamic and static comparison tests were carried out and a rapid test method for subgrade bearing capacity based on Soil Stiffness Gauge (SSG) and Portable Falling Weight Deflectometer (PFWD) with dynamic modulus was proposed. To establish the connectivity of the static and dynamic modulus, modulus-dependent prediction model was developed. The results show that the multiplier power model with modulus ($$$$) is superior to the usual linear model (E_BB = 0.9415E_PFWD − 6.1507, R² = 0.9639; E_BB = 0.7878E_SSG − 0.4566,  R² = 0.8894), and it can replace the traditional BB method. PFWD and SSG were found to be reliable devices with faster and more accurate monitoring of the modulus change of the fine-grained soil subgrade. But they have the property of overestimating material modulus, the modulus ranking of the three instruments is obtained. In this way, it provides a reference for the dynamic and accurate determination and scientific evaluation of the bearing capacity of the highway subgrade.

The deformation of the main girder is an important manifestation of the overall stiffness of suspension bridges, which is essential for assessing bridge performance. Nevertheless, it is difficult to achieve satisfied prediction without fully considering the overall operational loads. To this end, this paper proposes a method to predict the deflection considering multiple operational loads using the monitoring data of a high-speed rail-cum-road suspension bridge. Initially, an improved Gaussian process regression (GPR) model utilizing Bayesian optimization was employed to predict the deformation of the main girder under the condition of nontrain loads. Furthermore, the distinct contributions of temperature, wind, and vehicle load were analyzed. Subsequently, based on the strain and deflection induced by train loads, the sum of sinusoids method was proposed to construct fitting and shape function (FSF) for predicting the main girder deformation under the influence of train loads. Ultimately, the deformation considering overall loads was obtained by adding the deformation under the nontrain and train loads, and the predicted deformation result was verified using the measured data. When compared to other state-of-the-art machine learning algorithms, namely, artificial neural network (ANN), support vector machine (SVM), and decision tree (DT), the improved GPR demonstrates the highest accuracy in predicting the deformation of the main girder under nontrain loads with R² of 0.9478. In addition, the proposed sum of sinusoids FSF method accurately predicted the deformation of the main girder caused by train loads, with R² of 0.934. The deformation of the main girder under the influence of overall loads can lay a foundation for the early warning and evaluation of the suspension bridges.