Structural Control & Health Monitoring最新文献

Structural health monitoring is essential for the lifecycle maintenance of tunnel infrastructure. Distributed fiber-optic sensor (DFOS) technology, which is capable of distributed strain measurement and long-range sensing, is an ideal nondestructive testing (NDT) approach for monitoring linear infrastructures. This research aims to develop a distributed sensing network utilizing DFOS for structural integrity assessment of concrete immersed tunnels. The primary innovations of this study lie in the development of a general flowchart for establishing a sensing network and obtaining reliable field data, as well as its subsequent validation through a detailed case study. Concentrated joint deformations in typical immersed tunnels, detectable by the DFOS, are key indicators of structural integrity. This study addresses crucial elements of field monitoring system design, including the selection of appropriate optical fibers or cables and the determination of vital interrogator system parameters. It also covers sensor parameter determination, installation techniques, field data collection, and postanalysis. Furthermore, this research is exemplified by a case study that illustrates the successful implementation of a distributed sensing network in an operational immersed tunnel, and monitoring data reveals cyclic structural deformations under impacts of daily tide and seasonal temperature variations. The data obtained from this network play a significant role in subsequent condition assessments of tunnel structures. The research findings contribute to the assessment of large-scale infrastructure health conditions through the application of DFOS monitoring.

In this study, a novel delamination detection method for composite materials is proposed through the innovative use of You Only Look Once v8 (YOLOv8), vibration analysis, and 2D continuous wavelet transform techniques. The method detects the location and size of damage more accurately than existing methods and avoids manual intervention in the detection process. Damage detection performed on the simulation dataset shows that the method is able to accurately identify the delamination location with IoU = 0.9906 and an average accuracy of 91.32%. The proposed method is then compared with the widely used YOLOv5 model, and the superior performance of the YOLOv8 model is verified, with a 37.93% improvement in training speed and 0.81% improvement in detection accuracy. In addition, an experimental dataset of four composite laminates with delamination damage is constructed. By using transfer learning, the performance of the pretrained network achieves a good precision up to 1. The method proposed in this study expands the range of tasks that can be accomplished by mode shape analysis and is very effective in real experiments.

In structural health monitoring (SHM), an important issue is the limited availability of measurement data due to the spatial sparsity of sensors installed on the structure. These measurements are insufficient to accurately depict the actual dynamic behavior and response of the structure. Therefore, full-field (i.e., every degree of freedom) structural response reconstruction based on sparse measured data has drawn a lot of attention in recent years. Kalman filter (KF) is an effective technology for response reconstruction (also known as state estimation), providing an optimal solution for systems that can be well-represented by a fully known Gaussian linear state-space model. This implies that both the process noise and measurement noise follow known zero-mean Gaussian distribution, which is impractical in many civil engineering applications considering the unavoidable modeling errors and variations of environmental conditions. To address this challenge, a data-physics hybrid-driven method, i.e., KalmanNet, is proposed in this study for response reconstruction of partially known systems. By integrating a recurrent neural network (RNN) module into the KF framework, KalmanNet can efficiently learn and compute the Kalman gain using available monitoring data, without any Gaussian assumptions or explicit noise covariance specifications (e.g., covariance matrices of process and measurement noise). Both numerical and experimental investigations are conducted to validate this method. The results demonstrate that under the influence of non-Gaussian noise and modeling errors, KalmanNet can effectively and accurately reconstruct the structural response from sparse measurements in real-time and has higher accuracy and robustness compared to traditional KF even with optimal parameter settings.

The geomechanical model test (GMT), a means of intuitively exploring the model’s failure modes and revealing failure mechanisms, is considered an effective approach for studying the structural characteristics of dams under complex geological conditions. However, during the overloading process of the model, the catastrophe trends of monitoring data are unclear, and catastrophe points differ at different monitoring sites. These factors have led to large errors in the judgment of researchers regarding the model’s state and misperception of the structural properties during the damage process. In this study, a comprehensive evaluation method for the model’s state intervals in the damage process is proposed. The criterion employed an interval analysis hierarchy process that considered the differences, consistency, and credibility (CDC-IAHP) among multiple decision-makers (DMs), effectively reducing the subjectivity of their judgments. Additionally, this process was combined with cusp catastrophe theory (CCT) to determine whether the model underwent an abrupt change at various overload factors comprehensively. This is the first time that CDC-IAHP and CCT have been combined as criterion for a comprehensive method on the damage process of concrete dams in GMTs, and was applied to the Wudu gravity dam, indicating its applicability is very good. Compared to the researcher’s judgment, this approach is used to analyze and judge the structural state more accurately and scientifically while reducing subjectivity, which can help to better understand the structural characteristics and bearing capacity of actual engineering projects.

Changing environmental conditions and measured noises often affect the dynamic responses of structures and can obscure subtle changes in the vibration characteristics caused by damage. To address this issue, a new method for classifying damage in offshore structures under varying environmental conditions and measured noises is proposed using linear discrimination analysis (LDA). Two sets of data on dynamic characteristics, one from healthy structures and the other from unknown testing structures, are used to determine the optimal projection vector. This vector is perpendicular to the discriminant hyperplane and is used for damage classification. The damage-sensitive features are extracted by projecting both sets of data onto this vector. These features are then used with the hypothesis test technique to determine the condition state of the testing structure. Numerical studies on offshore wind turbine structures and experimental validations of a deep-sea mining system are being conducted to evaluate the effectiveness of the proposed approach. The study also examines the impact of mode combinations, measured noises and samples on the performance of the approach. The results indicate that the proposed approach can accurately assess the structural health state even in the presence of environmental changes and noise contamination, even with limited samples. The promising performance of the approach will facilitate the establishment of an online structural monitoring system to ensure the safety of offshore structures.

Long-term and accurate monitoring of suspender cable tensions is particularly important for safe evaluation of cable suspension bridges or tied-arch bridges. Torsional deformation, commonly present in wire rope suspender cables (WR cables) during tensioning construction or in-service, has not been considered in the elasto-magneto-electric (EME) sensor system. This study investigated the effects of torsion on tension measurement and proposed an intelligent correction method without measuring the torsion angles per unit length. A calibration platform for full-scale WR cable is established with a rotation angle fixing device. Tension calibration experiments were carried out under free rotation condition without activating the angle fixing device and under various fixed rotation conditions by setting a series of initial fixed angles at the anchor head. It was found that the relative error for the EME sensor using the traditional calibration method under the free rotation condition could reach 11.72%. To improve the accuracy, an intelligent tension correction method for the torsion effect is proposed, which uses the experimental signals in various fixed conditions and the backpropagation neural network with K-fold cross-validation. The parameters of the BPNN were optimized by genetic algorithm, and it was found that the maximum relative error decreases from 11.72% to 5.24% and the maximum absolute error decreases from 21.75 kN to 14.67 kN for the condition of free rotation. Finally, the EME sensor with intelligent tension correction method was applied to a real suspension bridge. The measurement relative error of the field test decreases from 6.60% without the torsion compensation to 2.80% with the torsion compensation, which indicate that the proposed intelligent tension correction method can ensure the accurate tension measurement of the WR cables by the EME sensor.

This paper introduces a damage diagnosis method based on the reassignment method and matching networks (MNs) to study the structural health monitoring of aerospace composite material components. This aims to facilitate the mapping of signal features to complex failure modes. We introduce a signal processing technique based on the reassignment method, employing a sliding analysis window to re-estimate local instantaneous frequency and group delay. By utilizing the short-time phase spectrum of the signal, we correct the nominal time and frequency coordinates of the spectrum data, aligning them more accurately with the true support region of the analyzed signal. Subsequently, this paper developed a deep matching network (DMN) damage diagnosis model based on MNs. This model utilizes a convolutional neural network (CNN) to extract damage-related features from the signal and introduces the full context embedding (FCE) method to enhance the compatibility of sample embeddings. In this process, the embeddings of each sample in the training set should be mutually independent, while the embeddings of test samples should be regulated by the distribution of training set sample data. Ultimately, the damage category of test samples is determined based on cosine similarity. We validate our model using damage sample data collected from experiments and simulations conducted under varying components and operating conditions. Comparative assessments with five mainstream methods reveal an average accuracy exceeding 96%. This underscores the exceptional recognition accuracy and generalization performance of our proposed method in cross-operating condition fault diagnosis experiments concerning aircraft composite material components.

As the combination of springs and vertical stabilizers, the movable downward vertical central stabilizer (MDVCS) is proposed to further control the nonlinear flutter of super long-span suspension bridges in this paper. A series of flutter suppression tests of closed-box girders and twin-box steel girders with various MDVCSs are conducted. Based on the coupled flutter theoretical method, the sensitivity analysis of two important parameters including the height and stiffness of MDVCS are carried out to compare their nonlinear flutter control mechanism. The results show that the flutter critical wind speed (U_cr) of the closed-box girder continued to increase with the decrease of the height of the DVCS and the increase of spring stiffness, whereas the U_cr of the twin-box girder increased at first and then decreased. The cubic polynomial function and quadratic Holliday function are suitable to modify the correction coefficients of U_cr for the closed-box girder with various stiffnesses and heights of MDVCS, while the Lorentz peak-value function and cubic polynomial function are suitable to modify the U_cr of the twin-box girder. Furthermore, the MDVCS significantly changes the rules of two positive and negative aerodynamic damping ratios for the closed-box girder and two negative aerodynamic damping ratios for the twin-box girder. Besides, the peak vertical displacement amplitudes of the box girder are about half of the MDVCS, since both the height and stiffness of MDVCS alter the elliptical radius of vertical phase planes to affect the limit cycle oscillation of soft flutter, especially for the leeward MDVCS for the twin-box girder.

This paper proposes a novel reliable model predictive control (MPC) method for active vibration control of structure with nonprobabilistic uncertainties. First, the framework of reliable MPC is established by integrating nonprobabilistic reliability constraints into nominal MPC. Based on the first-order Taylor expansion and first-passage theory, an efficient nonprobabilistic reliability analysis method that is suitable for online computation is proposed. A nonprobabilistic Kalman filter is further proposed for determine system states and their uncertain region. Unlike most robust MPC approaches, the proposed reliable MPC focuses on the satisfaction of state constraints in terms of structural reliability and is more suitable for structures with stringent safety requirements. Compared to existing reliability-based vibration control methods, reliable MPC requires no knowledge of disturbance and exhibits greater adaptability to load environments. The effectiveness and superiority of the proposed reliable MPC are validated through a numerical example and an engineering case study.

In response to the data anomalies and frequent false alarms caused by harsh environments in long-term structural health monitoring (SHM), this study has reframed the detection of abnormal vibration data as a time series classification problem. This approach identifies multiple anomalous features, thereby reducing manual detection costs. The novel developed Convolutional Neural Network with Squeeze-and-Excitation and Multi-Head Self-Attention (CNN–SE–MHSA) employs a deep residual network structure with channel and spatial attention mechanisms, effectively handling the global long-term dependencies required for anomaly feature learning. It better understands and utilizes feature information across different levels and dimensions, enhancing classification accuracy in complex anomaly situations. Through t-SNE dimensionality reduction visualization and interpretability analysis, it is demonstrated that the model excels in identifying critical features. Furthermore, by generating simulated data with a variational autoencoder (VAE) and implementing transfer learning strategies based on these data, the issue of low recognition accuracy for complex anomaly data due to data imbalance can be effectively mitigated. In a 25-day long-term monitoring experiment of indoor tunnel lining structures, this method demonstrated an average accuracy rate exceeding 96% and a rapid detection capability within 16 min. The results indicate that this method achieves high accuracy in anomaly detection for long-term monitoring data, even when relying exclusively on time-domain data.