Yonghe Bridge Research Articles

Optimal tuning of three deep learning methods with signal processing and anomaly detection for multi-class damage detection of a large-scale bridge

Long-span bridges play a crucial role in urbanization, connecting communities across vast obstacles. Structural health monitoring techniques have been deployed on these bridges, generate large amounts of data through sensor measurements, requiring data-driven approaches like deep learning (DL) for effective analysis. However, feature extraction from time-domain vibration response signals poses challenges for DL methods. To address this, the study proposes utilizing signal processing techniques such as the multivariate empirical mode decomposition (MEMD) and Wavelet transform (WT) to extract essential features for damage classification. The incorporation of MEMD and WT aims to overcome limitations and process nonstationary and nonlinear signals effectively. Three DL techniques, long-short-term memory (LSTM), one dimensional convolutional neural network (1D-CNN), multi-layer perceptron (MLP) are tuned and applied to Structural Health Monitoring of Tianjin Yonghe Bridge (located in China) as a real-world case study, in order to detect its condition by Deep signal anomaly detection and identify types of the damage. A powerful meta-heuristic algorithm called Observer-Teacher-Learner-Based Optimization, is used to optimize both hyperparameters and architecture of each DL models. The results demonstrate that the optimally tuned DLs are successful in identifying types of damage, as well as the condition of the structure, for the Tianjin Yonghe Bridge. The average accuracy values are obtained as 98.13, 97.96, and 97.79 for 1D-CNN, LSTM, and MLP, respectively. Such optimally tuned DLs are evaluated as effective solutions for detecting damage on large-scale bridges by extracting statistical time-domain and time–frequency domain features using the WT and MEMD.

Unsupervised structural damage assessment method based on response correlations

Structural damage assessment for civil engineering structures has become an important research topic and has gained much attention for decades in the field of structural health monitoring (SHM). Current unsupervised structural damage assessment methods can achieve damage detection for practical civil engineering structures. However, these methods cannot fully mine the response correlations among sensors, thus are difficult to localize damage in practical engineering applications. Therefore, this study proposes an unsupervised structural damage assessment method with bidirectional long short-term memory (BiLSTM) networks to establish the response correlations among sensors. Numerical studies of a beam-like bridge model, an experimental study of Yellow Frame, and an in-field study of Yonghe Bridge were conducted to validate the accuracy and reliability of the proposed approach. The results show that unsupervised structural damage identification and localization can be achieved even in the case of sparse sensor deployment due to the fact that the response correlations among sensors can be well extracted. The proposed approach can be applied to damage identification of practical engineering structures under operational and ambient excitations and can automatically provide damage location warnings to ensure structural health and prevent catastrophic events.

A Novel Damage Detection Approach Based on Feature Extraction and Selection Using Machine Learning Without Signal Processing: A Case Study on the Tianjin Yonghe Bridge

A Novel Damage Detection Approach Based on Feature Extraction and Selection Using Machine Learning Without Signal Processing: A Case Study on the Tianjin Yonghe Bridge

A fast and efficient feature extraction methodology for structural damage localization based on raw acceleration measurements

Many damage detection strategies have been developed within the field of structural health monitoring, showing promising results in real-world applications. Most of them rely on the use of modal parameters to locate damage in structures. However, they are highly dependent on the process of extracting such characteristics, which contains uncertainties and needs to be adapted specifically for each structure. Recently, a special focus is being given to techniques based only on the use of raw structural acceleration measurements due to their relatively lower computational complexity manipulation. Thus, this work proposes an efficient feature extraction methodology for structural damage localization based on raw vibration signals. Furthermore, its ability to indicate damage quantification is also investigated. The proposed approach consists in extracting sensitive features from time, frequency, and quefrency domains. Then, percentile intervals are defined for each feature regarding the structures' healthy state. Finally, a damage index is estimated based on the number of outlier features of the structures' damaged state. In this sense, the sensor with the highest number of outlier features indicates the damage location. To assess this methodology, four different applications are studied: a numerical two-dimensional frame, a simply supported beam tested in laboratory, a three-dimensional frame also tested in laboratory, and the Tianjin Yonghe Bridge in China. Results show that the proposed approach is not only able to correctly indicate damage locations but also to give insights about their magnitudes.

Generative Adversarial Network for Damage Identification in Civil Structures

In recent years, many efforts have been made to develop efficient deep‐learning‐based structural health monitoring (SHM) methods. Most of the proposed methods employ supervised algorithms that require data from different damaged states of a structure in order to monitor its health conditions. As such data are not usually available for real civil structures, using supervised algorithms for the health monitoring of these structures might be impracticable. This paper presents a novel two‐stage technique based on generative adversarial networks (GANs) for unsupervised SHM and damage identification. In the first stage, a deep convolutional GAN (DCGAN) is used to detect and quantify structural damages; the detected damages are then localized in the second stage using a conditional GAN (CGAN). Raw acceleration signals from a monitored structure are used for this purpose, and the networks are trained by only the intact state data of the structure. The proposed method is validated through applications on the numerical model of a bridge health monitoring (BHM) benchmark structure, an experimental steel structure located at Qatar University, and the full‐scale Tianjin Yonghe Bridge.

Unsupervised Structural Damage Detection Technique Based on a Deep Convolutional Autoencoder

Structural health monitoring (SHM) is a hot research topic with the main purpose of damage detection in a structure and assessing its health state. The major focus of SHM studies in recent years has been on developing vibration‐based damage detection algorithms and using machine learning, especially deep learning‐based approaches. Most of the deep learning‐based methods proposed for damage detection in civil structures are based on supervised algorithms that require data from the healthy state and different damaged states of the structure in the training phase. As it is not usually possible to collect data from damaged states of a large civil structure, using such algorithms for these structures may be impractical. This paper proposes a new unsupervised deep learning‐based method for structural damage detection based on convolutional autoencoders (CAEs). The main objective of the proposed method is to identify and quantify structural damage using a CAE network that employs raw vibration signals from the structure and is trained by the signals solely acquired from the healthy state of the structure. The CAE is chosen to take advantage of high feature extraction capability of convolution layers and at the same time use the advantages of an autoencoder as an unsupervised algorithm that does not need data from damaged states in the training phase. Applications on the two numerical models of IASC‐ASCE benchmark structure and a grid structure located at the University of Central Florida, as well as the full‐scale Tianjin Yonghe Bridge, prove the efficiency of the proposed algorithm in assessing the global health state of the structures and quantifying the damage.

Damage Detection of Tension Pendulums in Cable-Stayed Bridges Using Structural Frequency Variance

Structural health monitoring represents an efficient way to diagnose the condition of cable-stayed bridges. Most studies assessing the condition of cable-stayed bridges and detecting damage to them have been developed to monitor the longitudinal performance of bridge bearings. However, few studies have aimed to detect the damage to tension pendulums, which are usually tie-down cables that resist girder uplift from auxiliary piers. This study aims to develop an effective method for detecting the damage of tension pendulums in cable-stayed bridges using structural modal frequencies. First, the sensitivity of the bridge frequency to tension pendulum damage is derived according to the Rayleigh method. Then, a finite-element simulation of a cable-stayed bridge is conducted to validate the rationality of the frequency index. Finally, a case study on tension pendulum damage is carried out for the cable-stayed Yonghe Bridge. The results indicate that the frequency change is far more sensitive to the damage of the tension pendulums than to the damage of the stay cables. The failure of tie-down cables will result in a substantial decrease in the bridge frequency, up to approximately 35% for the first vertical bending mode of the bridge. In practice, the damage of tension pendulums can be effectively detected by a sudden decrease in the measured frequency of cable-stayed bridges.

Damage detection of a cable-stayed bridge based on combining effective intrinsic mode functions of empirical mode decomposition using the feature selection technique

This paper presents a novel approach for detecting damage in a cable-stayed bridge based on adaptive signal processing and feature extraction. The Empirical Mode Decomposition (EMD) technique is employed in this research to obtain the Intrinsic Mode Functions (IMFs). Several features in time and frequency domains are extracted from IMFs as damage-sensitive features. Next, to reduce false alarms, several spectral parameters are introduced as effective features. A Support Vector Machine (SVM) is employed as a classifier to compare the detection ability of the proposed and conventional features. The second part of this research employs the application of several feature selection techniques that are based on the filter, wrapper, and embedded approaches to obtain the most effective IMFs and then combine them to create a subset that carries the most significant information about the structural damage. To investigate the capability of the proposed approach, vibration acceleration data obtained from a real structure, the Yonghe bridge, is employed. Results indicate that the proposed approach can identify damage with high accuracy.

Replacement of Cable Stays in Yonghe Bridge in Tianjin, China

Cable replacement is one of the most important approaches for retrofitting an existing cable‐stayed bridge. The Yonghe Bridge opened to traffic in 1987 and was the largest concrete cable‐stayed bridge in Asia at the time. After nearly 20 years in service, its stay‐cable system was completely replaced. The present paper details the key technical points of cable replacement in this bridge, including causes of replacement, cable replacement design, construction monitoring during replacement, and cable tension tuning after replacement. Based on the measurements during construction, the variation in structural condition indexes was thoroughly investigated, such as cable forces throughout the bridge, geometry profile of bridge deck, and tower horizontal displacements. The structural changes due to cable replacement were evaluated, and then, a relatively desirable scheme of cable adjustment was correspondingly put forward. The measured results showed that, influenced by the accumulated tolerances of cable forces during construction, cable replacement may often lead to a further deviation of structural condition. In this instance, cable adjustment is an effective compensation measure to improve the overall structural condition of the bridge. The practice of cable replacement in the Yonghe Bridge has provided some valuable experience or beneficial references for the similar retrofitting cases.

Determining the number of measurements for compressive sensing of traffic-induced vibration data

Compressive Sensing (CS) is an emerging signal sampling technique, which can be useful in the long-term monitoring of civil infrastructures by reducing the storage space and transmission bandwidth drastically. In this paper, a fully deterministic approach is adopted to enhance the previously proposed CS-based methods. First, the sparsifying dictionary is trained using a vibration data set. Then, a deterministic projection matrix is computed based on the trained dictionary. Second, a new index is defined to determine the number of measurements in advance, without any trial and error in the reconstruction stage. This index which is coined as NPI (Normalized Power Index) is derived using the singular value decomposition of the trained dictionary. The capability of the proposed method is investigated using vibration signals of the Tianjin Yonghe Bridge with traffic excitation. Both accuracy and computational time of the deterministic CS was compared to different data compression algorithms.

Damage detection of a cable-stayed bridge using feature extraction and selection methods

In this paper, a novel approach is presented to detect damage in a cable-stayed bridge based on feature extraction and selection. In the first part of the paper, several features are used in time domain, frequency domain, and time-frequency domain to detect damage. Next, several spectral parameters are introduced as effective features to reduce false alarms. Support vector machine (SVM) technique is employed as a classifier to compare the detection ability of the proposed and conventional features. In the second part of the paper, several feature selection techniques including forward selection, backward elimination, the minimum redundancy and max relevancy, ReliefF and SVM approach based on recursive feature elimination (SVM-RFE) algorithm are employed to improve feature extraction accuracy through selecting and ranking the most discriminative feature as an optimal feature subset. The capability of the proposed approach to detect damage is investigated by a data set obtained from a health monitoring system of a real benchmark problem. The data consists of vibration acceleration of the Yonghe bridge in both healthy and damaged states. Results show that the proposed procedure can effectively reduce false alarms.

Fractal dimension based damage identification incorporating multi-task sparse Bayesian learning

Sensitivity to damage and robustness to noise are critical requirements for the effectiveness of structural damage detection. In this study, a two-stage damage identification method based on the fractal dimension analysis and multi-task Bayesian learning is presented. The Higuchi’s fractal dimension (HFD) based damage index is first proposed, directly examining the time-frequency characteristic of local free vibration data of structures based on the irregularity sensitivity and noise robustness analysis of HFD. Katz’s fractal dimension is then presented to analyze the abrupt irregularity change of the spatial curve of the displacement mode shape along the structure. At the second stage, the multi-task sparse Bayesian learning technique is employed to infer the final damage localization vector, which borrow the dependent strength of the two fractal dimension based damage indication information and also incorporate the prior knowledge that structural damage occurs at a limited number of locations in a structure in the absence of its collapse. To validate the capability of the proposed method, a steel beam and a bridge, named Yonghe Bridge, are analyzed as illustrative examples. The damage identification results demonstrate that the proposed method is capable of localizing single and multiple damages regardless of its severity, and show superior robustness under heavy noise as well.

Blind Modal Identification Using Limited Sensors through Modified Sparse Component Analysis by Time‐Frequency Method

AbstractOutput‐only modal identification with limited sensors available may be encountered in practical applications. Sparse component analysis (SCA), which is one of the most popular methods among blind source separation (BSS) techniques, is known as its capability for handling underdetermined BSS problems and nonstationary signals. Single‐source‐points (SSPs) where only one mode makes contribution are adopted to estimate the modal matrix. However, the accuracy of SSP detection is influenced by the number of sensors, because the detecting condition for SSPs may be satisfactory for multiple‐source‐points (MSPs). Besides, the estimation of frequencies and damping ratios is based on the sparse reconstruction of time‐frequency domain modal responses, where the number of sensors will have an impact on the accuracy. To improve the performance of blind modal identification using limited sensors, a method which is called SCA‐TF by modifying SCA through time‐frequency method is proposed in this paper. The TF representation in SCA is employed to modify the SSP detection process and identify frequencies and damping ratios instead of recovering the modal responses. In order to improve the accuracy of modal matrix estimation, the result of SSP detection is used as a preliminary for further refinement. Then the false SSPs are removed from the preliminary result through the power spectra which are constructed by short‐time Fourier transform coefficients. Subsequently, the estimation of modal matrix is accomplished by clustering the refined SSPs. Frequencies and damping ratios are extracted directly from the generalized spectra matrix which is formed by the short‐time Fourier transform coefficients. Because the identification of modal parameters is implemented directly in time‐frequency domain, the reconstruction of modal responses is omitted. The results of the numerical studies show that SCA‐TF can identify the modal parameters accurately. To investigate the performance of SCA‐TF in practical structures, the acceleration measurements from Tianjin Yonghe Bridge are analyzed to identify the modal parameters. The results of the Yonghe Bridge demonstrate the effectiveness of the proposed method to perform blind modal identification in practical structures.

Damage identification of a large cable-stayed bridge with novel cointegrated Kalman filter method under changing environments

Damage identification is an indispensable part for successive applications of structural health monitoring. In practical applications, however, time-varying environmental and operational conditions, such as temperature and external loadings, often overwhelm the subtle structural changes caused by damage. It is therefore of great significance to remove those structural changes (damage features) caused by external influences from actual structural damage. In this paper, a new damage identification method based on Kalman filter and cointegration (KFC) is developed, and the environmental effects on damage indicator are removed by the cointegration process of the Kalman filtered coefficients. The cointegration relationship between structural frequencies is first established with augmented Dickey–Fuller test and Johansen procedure. The cointegration coefficients are then used to constitute the Kalman filter (KF) state vector, and the recursive KF process is intrigued to on-line estimate the change of structure states. To enhance the importance of incoming new observations in the KF, we introduce an adaptive fading factor into the conventional KF. Numerical simulation of a truss bridge is used to validate the effectiveness of the proposed KFC method for damage identification under varying temperature, even with 10% noise. Finally, the KFC method is applied to a cable-stayed bridge built in China (Tianjin Yonghe Bridge), and two structural damage scenarios are successfully identified. The advantages of the proposed KFC method are its ability to eliminate ambient temperature influences and identify structural damage on-line.

The Second Maintenance and Strengthening of Yonghe Bridge

Tension rocker bearing (TRB) cables installed at auxiliary piers were critical members to sustain the structural system in a prestressed concrete cable-stayed bridge. Once these cables fractured or broken, its structural system would be transferred. Based on diseases caused by fracture of TRB cables in Yonghe Bridge, corresponding maintenance or strengthening measures were described in detail. These measures included replacement of TRB cables, strengthening of the closure segment of main girder at each side span, and adjustment of stay cable forces. Monitoring results showed that, its structural system was finally and completely rehabilitated, and the safety factor of new TRB cables is enough under the most unfavorable load combination. Moreover, the line shape of main girder and the inclined displacement at the top of each pylon were improved effectively. Thus, Yonghe Bridge accumulated some valuable experience for maintenance or strengthening of existing prestressed concrete cable-stayed bridges, and also made a useful exploration.

Analysis the Cable Replacement Method of Tianjin Yonghe Bridge

Taking Tianjin Yonghe bridge for example, the finite element method is adopted to simulate each stages of cable replacement process, the dates of tension, the alignment of main beam and the change of stress before and after cable replacement are compared and analysed, The measured value is greatly consistent with calculated value, Stress state of the bridge has been greatly improved, the weight of main beam bearing by the cable will be shared by a few closed cable after unloading, so the cables which are more serious corroded should be replaced.

A Method for Seismic Capacity Evaluation of CFST Arch Bridges

Based on strength and ductility, the seismic capabilities of the chords of arch ribs, web members, horizontal integrations, suspenders and beams of suspenders of concrete filled steel tube (CFST) arch bridges are studied, then the seismic capability of the bridge is obtained. Firstly, the internal forces of the members are calculated respectively by finite element under the actions of gravity representative value and small earthquakes (0.05g). Then the ultimate bearing capacities, the ratio of ductility and the reduction coefficients of earthquake are obtained according to the failure modes. Finally, yield accelerations of ground movement are multiplied by the reduction coefficients of earthquake, which the resistance seismic capabilities of the members Ac are obtained. The seismic capability of bridge is the minimum Ac. Taking Nanning Yonghe Bridge that is a 346m CFST arch bridge as example, the seismic capability is evaluated that the bridge can resisted the earthquake which the acceleration of ground movement is respectively 0.677g. The results show that the method of seismic capability evaluation is a feasible and efficient method for seismic capability evaluation of CFST arch bridge.

Sensitivity and analysis GPS signals based bridge damage using GPS observations and wavelet transform

This paper uses GPS monitoring system on the Yonghe Bridge. After a year from re-construction of this bridge, cracks were pronounced 48.2 m far from the beginning of the abutment. To analyze the movements of bridge tower, the current operational safety, the cause of bridge cracks and sensitivity of GPS signals, and its movements were observed under different stress factors such as wind speed, temperature change and traffic loads. The observed lateral, longitudinal of bridge’s towers were studied. Two analytical methods namely; DWT and FT were used for the analysis of observation data. The analysis of test results indicates that: (1) The STFT is a significant step forward from the traditional FFT in terms of structural response analysis. (2) The sensitivity of GPS signals does not depend on the position of GPS antenna; and (3) the cracks of bridge deck bring out the frequency of the bridge tower movement transient characteristics.

Simple geometrical model to analyze the motion detection of bridges based-GPS technique: case study Yonghe Bridge

This study deals with the viability of using a designed geometrical model consists of plane, polar coordinates (PC) and span length in the determination of bridges deformation. The data of a Tianjin Yonghe bridge located in the southern part of China as collected by RTK-DGPS technique and Accelerometer were used in the analysis. Kalman filter and fast Fourier transformation (FFT) analyses were used to determine the frequency. The results indicate that the designed plane and PC geometrical model are easy to calculate the long-time structural deformation monitoring. In addition, the observed frequency using GPS with the rate of 20 Hz doesn\'t give correction natural frequency of the observation structures.

Fatigue Cumulative Damage Evaluation Based On Smart Stay Cables

Abstract Stay cables are one of the most critical structural components of cable-stayed bridge. However, stay cables readily suffer from fatigue damage, corrosion damage and their coupled effects. Thus, health monitoring of stay cables is important for ensuring the integrity and safety of a bridge. The application of the smart stay cables on the Tianjin Yonghe Bridge is demonstrated and the vehicle fatigue load effects smart stay cables are evaluated based on field monitoring data. Secondly, based on fatigue test results of corroded wires obtained from dissection of actual parallel wire cables which were used on Tianjin Yonghe cable-stayed bridge, the fatigue properties of corroded cable are investigated by the method of Monte Cazlo simulation. The results of fatigue lives and corrosion degree of corroded wire are presented. Comparisons between the original design information and fatigue test results, it can be seen that corrosions make the fatigue lives of wires decreasing sharply and the cable fatigue l...