Application of unsupervised support vector machine for condition assessment of concrete structures

The main objective of structural health monitoring is to provide reliable information about the health state of the critical structures by implementing a damage characterization strategy to detect the presence of damage, location, severity as possibly failure prediction as soon as the damage occurs. This paper presents a robust approach to detect and characterize a gradually evolving damage based on time responses data captured from a steel reinforced concrete structure. The presented method is in the context of unsupervised and non-model-based approaches, hence, there is no need for any representative numerical/finite element model of the structure to be built. In this work, we propose one-class support vector machine as an anomaly detection method. One-class support vector machine fits well for damage diagnosis in structural health monitoring since there may exist many damaged patterns and one-class support vector machine can detect all of them as anomalies. To demonstrate the feasibility of the method in the detection and assessment of a gradually evolving deterioration, a test bed was established to replicate a concrete jack arch which is a main structural component on the Sydney Harbour Bridge – one of Australia’s iconic structures, see Figure 1. The structure is a concrete cantilever beam with an arch section which is located on the eastern side of the bridge underneath the bus lane. It is assumed that the structure is subjected to Gaussian white noise excitation. The response of the structure was collected by 10 uniaxial accelerometers placed in frontal face of the jack arch. The measurement was conducted for 2 seconds at a sampling rate of 8 kHz, resulting in 16000 samples for each event. Figure 1 Illustration of the test bed. https://s3-euw1-ap-pe-df-pch-content-public-p.s3.eu-west-1.amazonaws.com/9781315207681/cd556cd4-4dcf-4efe-8e29-56fc67b8bfbd/content/fig321_1.jpg/> A crack is introduced in the structure using a cutting saw and its length is progressively increased in four stages while the depth was constant; these four damage cases correspond to less than 0.5% reduction in the first three modes of the structure. Figure 2 presents the average of all decision values obtained from all sensors for healthy state and different damage states using one-class support vector machine. As seen, the method is able to successfully separate the healthy state from the damage states and as the severity of damage increases, higher negative decision values are obtained. Figure 2 Damage detection using decision values of all test events. https://s3-euw1-ap-pe-df-pch-content-public-p.s3.eu-west-1.amazonaws.com/9781315207681/cd556cd4-4dcf-4efe-8e29-56fc67b8bfbd/content/fig321_2.tif/>

Structural health monitoring has been increasingly used due to the advances in sensing technology and data analysis, facilitating the shift from time-based to condition-based maintenance. This work is part of the efforts which have applied structural health monitoring to the Sydney Harbour Bridge – one of Australia’s iconic structures. It combines dimensionality reduction and pattern recognition techniques to accurately and efficiently distinguish faulty components from well-functioning ones. Specifically, random projection is used for dimensionality reduction on the vibration feature data. Then, healthy and damaged patterns of bridge components are learned in the lower dimensional projected space using supervised and unsupervised machine learning methods, namely, support vector machine and one-class support vector machine. The experimental results using data from a laboratory-based building structure and the Sydney Harbour Bridge showed high feasibility of applying machine learning techniques to dimensionality reduction and damage detection in structural health monitoring. Random projection combined with support vector machine significantly reduces the computational time while maintaining the detection accuracy. The proposed method also outperformed popular dimensionality reduction techniques. The computational time of the method using random projection can be more than 200 times faster than that without using dimensionality reduction while still achieving similar detection accuracy.

Development of Correlation Between Ultrasonic Pulse Velocity and Rebound Hammer Test Results for Condition Assessment of Concrete Structures for Sustainable Infrastructure Development

This research focused on finding the structural condition of a precast parking deck using a GPR. A GSSI SIR 4000 Bridge Scanner GPR was deployed for corrosiveness mapping of a precast deck. Small concrete slabs were prepared in the laboratory with fresh and corroded rebars to identify their differences in GPR scanned images. The GPR response from the slab indicated that the reflection amplitude obtained from the scan of the target rebar decreases with increasing depth. GPR B-scans from the fresh and corroded rebar indicated increase in corrosion. The field experiment was performed following the standard bridge deck survey procedures. Both numerical and image-based analysis methods were followed for developing the corrosiveness maps of the precast deck section. The corrosiveness map showed a few critical sections in the precast deck. A visual inspection of the precast deck matched the fault locations in the corrosiveness map. The present condition of a precast deck section was found satisfactory except for a few areas of serious cracking that need repairs to control the further growth of corrosion. GPR results showed higher accuracy with the field inspection outcomes demonstrating that a GPR can be used as a reliable non-destructive tool for the condition assessment of concrete structures.

Quality assurance and condition assessment of concrete structures is an important topic world-wide due to the aging infrastructure and increasing traffic demands. Common topics include, but are not limited to, localisation of rebar or tendon ducts, geometrical irregularities, cracks, voids, honeycombing or other flaws. Non-destructive techniques such as ultrasound or radar have found regular, successful practical application but sometimes suffer from limited resolution and accuracy, imaging artefacts or restrictions in detecting certain features. Until the 1980s X-ray transmission was used in case of special demands and showed a much better resolution than other NDT techniques. However, due to safety concerns and cost issues, this method is almost never used anymore. Muon tomography has received much attention recently. Novel detectors for cosmic muons and tomographic imaging algorithms have opened up new fields of application, such as the investigation of freight containers. Muon imaging also has the potential to fill some of the gaps currently existing in concrete NDT. As a first step towards practical use and as a proof of concept we used an existing system to image the interior of a reference reinforced 600 kg concrete block. Even with a yet not optimized setup for this kind of investigation, the muon imaging results are at least of similar quality compared to ultrasonic and radar imaging, potentially even better. The data acquisition takes more time and signals contain more noise, but the images allowed to detect the same important features that are visible in conventional high energy X-ray tomography. In our experiment, we have shown that muon imaging has potential for concrete inspection. The next steps include the development of mobile detectors and optimising acquisition and imaging parameters.

Recent failures of bridges worldwide highlight the need for robust structural health surveillance systems to provide alerts for early damage and prevent severe loss. Structural health monitoring (SHM) continues to evolve and gain wide adoption among asset managers to mitigate catastrophic failure risks of structural systems. With the prolific integration of SHM with artificial intelligence and internet of things systems becoming prevalent, several data-driven approaches have gained popularity in SHM. However, it is challenging to obtain ground-truth-labelled damage samples for supervised machine learning methods. A class of shallow machine learning methods known as unsupervised methods (e.g. the one-class support vector machine (OCSVM)), which only uses unlabelled data from a healthy state, has been shown to achieve promising results. However, shallow machine learning methods require laborious expert-crafted feature inputs for optimal performance. Inspired by the automatic feature-extraction ability of deep neural networks, unsupervised methods (deep learning models trained without supervisory ground truth labels) such as autoencoder (AE)-based methods were explored to detect structural damage. Results obtained on both real-world data from Sydney Harbour Bridge, one of Australia's iconic structures, and data collected from laboratory specimens demonstrate the effectiveness of the AE-based method over conventional shallow methods such as the OCSVM.

Machine learning algorithms have been employed extensively in the area of structural health monitoring to compare new measurements with baselines to detect any structural change. One-class support vector machine (OCSVM) with Gaussian kernel function is a promising machine learning method which can learn only from one class data and then classify any new query samples. However, generalization performance of OCSVM is profoundly influenced by its Gaussian model parameter \(\sigma \). This paper proposes a new algorithm named Appropriate Distance to the Enclosing Surface (ADES) for tuning the Gaussian model parameter. The semantic idea of this algorithm is based on inspecting the spatial locations of the edge and interior samples, and their distances to the enclosing surface of OCSVM. The algorithm selects the optimal value of \(\sigma \) which generates a hyperplane that is maximally distant from the interior samples but close to the edge samples. The sets of interior and edge samples are identified using a hard margin linear support vector machine. The algorithm was successfully validated using sensing data collected from the Sydney Harbour Bridge, in addition to five public datasets. The designed ADES algorithm is an appropriate choice to identify the optimal value of \(\sigma \) for OCSVM especially in high dimensional datasets.

Structural Health Monitoring (SHM) is a condition-based maintenance technology using sensing systems. In SHM, the use of domain knowledge is essential: it motivates the use of machine learning approaches; it can be used to extract damage sensitive features and interpret the results by machine learning. This work focuses on two SHM problems: damage identification and substructure clustering. Our solutions to address them are based on machine learning techniques and robust feature extraction using domain knowledge. In the first problem, damage sensitive features were extracted using a frequency domain decomposition, followed by a robust one-class support vector machine for damage detection. In the second problem, a novel clustering technique and spectral moment feature were utilised for substructure grouping and anomaly detection. These methods were evaluated using data from lab-based structures and data collected from the Sydney Harbour Bridge. We obtained high damage detection accuracies and were able to assess damage severity. Furthermore, the clustering technique was able to group substructures of similar behaviour and detect spatial anomalies.

<p>Machine learning (ML) provides a promising paradigm for discovering and modeling structural performances and conditions through the deep mining of structural health monitoring data. This paper exhibits recent progress of ML-based structural health monitoring and condition assessment for long-span bridges. A series of novel algorithms for bridge condition assessment via correlation modeling between structural responses, computer vision-assisted structural damage detection, and data mining for wind effects are introduced. First, correlation modeling between different bridge responses is investigated by both the time series data and probability distribution, further assisting bridge condition assessment. Second, several novel CNN architectures and a few-shot meta-learning framework are also established for CV-assisted bridge damage detection. Third, wind-induced vibrations of the bridge in site are identified and modeled to predicate structural responses and evaluate operation conditions. Results show that ML techniques indeed improve the state of the art in structural health monitoring and condition assessment for long-span bridges.</p>

The changing landscape of concrete bridge infrastructure health monitoring and informed maintenance strategies: A decade of developments and trends

Foreword

Vibration monitoring and acoustic emission sensing during progressive load tests of corroded reinforced concrete beams

Structural Health Monitoring (SHM) based on Automated Operational Modal Analysis (A-OMA) has gained increasing importance in the conservation of heritage structures over recent decades. In this context, finite element model updating techniques using modal data constitute a commonly used approach for damage identification. Nevertheless, the large number of simulations usually involved in the associated minimization problem hinders the application to real-time condition assessment. This is especially critical for historic buildings, where the modelling of complex geometries involves large computational burdens. Alternatively, surrogate models offer an efficient solution to replace computationally demanding numerical models and so perform continuous model updating. In this light, this paper presents a surrogate-based model updating approach for online assessment of historic buildings and its application to a medieval masonry tower, the Sciri Tower in Perugia (Italy). Using modal properties identified by A-OMA, the proposed approach allows the continuous fitting of certain damage-sensitive parameters of the structure. To do so, three different surrogate models are considered, including the quadratic response surface method, Kriging, and Random Sampling High-Dimensional Model Representation, and their effectiveness is compared from an SHM perspective. The reported results demonstrate the suitability of the proposed methodology for tracking the temperature-dependent intrinsic properties of the tower.

In this study, the surface response to excitation method (SuRE) is investigated using a data-driven method for load monitoring on a laminated composite plate structure. The SuRE method is an emerging approach in ultrasonic wavebased structural health monitoring (SHM) field. In this method, a range of high-frequency, surface-guided waves are excited on the structure using piezoceramic elements. The waves propagate on the structure and interact with internal or surface damages. Initially, a baseline data of the intact structure is created by measuring the frequency transfer function between the excitation and sensing point. The integrity of structure is evaluated by monitoring changes in the frequency spectrums. The SuRE method has effectively been used for a variety of SHM applications including the detection of loose bolts, delamination in composite structures, internal corrosion in pipelines, and load and impact monitoring. Data obtained using the SuRE method was used for identifying the location of the applied load on a laminated composite plate using Support Vector Machine (SVM). A set of two piezoelectric elements were attached on the surface of the plate. A sweep excitation (150-250 kHz) generated surface-guided waves, and the transmitted waves were monitored at the sensory positions. The reference data set was measured simultaneously from the sensors. The plate was subjected to static loads while health monitoring data was being captured using the SuRE method. The confusion matrix indicated that the model classified correctly with up to 99.8% accuracy.

Machine learning techniques have great potential in structural health monitoring (SHM) and damage assessment of precast concrete (PC) structures. However, traditional damage assessment methods, such as finite element analysis, structural reliability analysis, and model testing, have limitations in terms of accuracy, efficiency, and applicability to complex real-world scenarios. This article proposes a hybrid machine learning approach combining random forest (RF) and support vector machine (SVM) for damage detection and classification in PC structures. The proposed RFSVM method utilizes continuous wavelet transform (CWT) to extract damage-sensitive features from vibration response data and employs RF for initial feature selection and damage classification. The RF output is then concatenated with the original features to form an enhanced feature vector, which is fed into the SVM model for precise damage type identification. The RF-SVM model is trained and tested using vibration response data generated from a simplified PC frame structure finite element model built in OpenSEES software. The experimental results demonstrate that the proposed method achieves an overall accuracy of 82.5% in detecting and classifying typical damage types in PC structures, with high precision, recall. However, the model's performance in identifying joint connection damage is relatively lower due to the subtle changes in vibration responses caused by this damage type. Compared to traditional methods that require extensive damage scenario data, the RF-SVM method simplifies the data preparation process and reduces computational complexity by utilizing only baseline data for training. This hybrid approach not only advances the field of SHM for PC structures but also offers a robust tool for infrastructure management, potentially increasing safety and reducing maintenance costs.