Search

Detecting anomalies in the vibrational features of age-old buildings is crucial within the Structural Health Monitoring (SHM) framework. The SHM techniques can leverage information from onsite measurements and environmental sources to identify the dynamic properties (such as the frequencies) of the monitored structure, searching for possible deviations or unusual behavior over time. The Temporal Fusion Transformer (TFT) network is a deep learning algorithm designed for multi-horizon time series forecasting and initially tested on electricity, traffic, retail, and volatility problems. In this paper, it is applied to SHM. More precisely, the TFT approach is adopted to investigate the behavior of the Guinigi Tower located in Lucca (Italy) and subjected to a long-term dynamic monitoring campaign. The TFT network is trained on the tower's experimental frequencies enriched with other environmental parameters. The transformer is then employed to predict the vibrational features (natural frequencies, root mean squares values of the velocity time series) and detect possible anomalies or unexpected events by inspecting how much the actual frequencies deviate from the predicted ones. The TFT technique is used to detect the effects of the Viareggio earthquake that occurred on 6 February 2022, and the structural damage induced by three simulated damage scenarios. [ABSTRACT FROM AUTHOR]

Structural Health Monitoring has the potential to enhance the safety and serviceability of our aging infrastructures by detecting anomalies at an early stage. Bayesian Dynamic Linear Models (BDLM) have been shown to be effective at detecting anomalies by extracting structural patterns and latent variables from complex and noisy time series. However, the autoregressive component modelling the stationary prediction errors in most BDLM has a tendency to wrongfully capture patterns that should be attributed to anomalies, and thus hinders their detectability. This paper proposes a new bounded autoregressive (B A R) component, which imposes constraints on the autoregressive latent process with a new mixture Rectified Linear activation Unit. The B A R component is probabilistically verified on synthetic data using a new F1t metric, and is validated using real observations collected on a bridge and on a dam located in Canada. The experimental results demonstrate that the B A R model surpasses the performance of the existing autoregressive component with (1) an improved accuracy at estimating hidden states, (2) an early detection of anomalies, (3) a capacity to detect smaller anomaly magnitudes, and (4) the ability to control the tradeoff between the anomaly detectability and the false alarm rate. [ABSTRACT FROM AUTHOR]

Data-driven machine-learning algorithms generally suffer from a lack of labelled health-state data, mainly those referring to damage conditions. To address such an issue, population-based structural health monitoring seeks to enrich the original dataset by transferring knowledge from a population of monitored structures. Within this context, this paper presents a transfer learning approach, based on domain adaptation, to leverage information from completely-labelled bridge structure data to accurately predict new instances of an unknown target domain. Since intrinsic structural differences may cause distribution shifts, domain adaptation attempts to minimise the distance between the domains and to learn a mapping within a shared feature space. Specifically, the methodology involves the long-term acquisition of natural frequencies from several structural scenarios. Such damage-sensitive features are then aligned via domain adaptation so that a machine-learning algorithm can effectively utilise the labelled source domain data and generalise well to the unlabelled target-domain data. The described procedure is applied to two case studies, including the Z24 and the S101 benchmark bridges and their finite element models, respectively. The results demonstrate the successful exchange of health-state labels to identify the damage class within a population of bridges equipped with SHM systems, showing potential to reduce computational efforts and to deal with scarce or poor data sets in application to bridge network monitoring. • A data-based domain adaptation framework is applied to bridge monitoring. • Domain Adaptation aligns and transform damage-sensitive features into a shared feature space. • Joint Domain Adaptation and Normal Condition Alignment are compared and discussed. • The KNN algorithm uses the transformed features to perform damage detection and classification. • The method is applied to transfer health-state labels across real bridges and finite element models. [ABSTRACT FROM AUTHOR]

Accurate condition monitoring has been a major challenge among railroad management authorities as they work to minimize collisions that lead to fatalities or damage to railroads infrastructure. Hence, research and technological developments in railroad maintenance and inspection are vital. Several research studies, on the inspection and defect detection techniques for railroads have been conducted by scholars. Despite the significant advancements made in this area, extensive studies are still required to enhance the accuracy of prognostic methods for railroad structural health monitoring (SHM) and condition monitoring (CM). Distributed acoustic sensing (DAS) has been recognized as a promising measurement method because of its quick sensing capabilities over long distances and for massive structures. DAS systems are classified according to the optical sensing quality and sensing range. As DAS produces large noisy datasets, in the case of railroad applications, algorithms for precise real-time and reliable analysis are essential. Meanwhile, data-driven and machine learning (ML) methods for defect detection have emerged as valuable approaches for SHM. Engineers and stakeholders can use ML algorithms to examine the large volumes of data produced by SHM systems to identify patterns and behaviors that might not appear in manual inspections. To support more precise and accurate maintenance and inspection for railroad systems, methodologies used to detect, identify, and characterize abnormal conditions in railroads using DAS and signal processing approaches for DAS large size and noisy signals, must be reviewed. Accordingly, in this literature survey, the applications of DAS methods for railroad CM are investigated. Among the variety of DAS methods, optical time domain reflectometry (OTDR) is reviewed in more details, since it is the most common approach in distributed fiber optic sensing. In addition, different OTDR-based DAS research for train tracking and railroad SHM are reviewed, and a comprehensive summary of different railroad defects is provided for further investigation. In all, this review paper provides a comprehensive background on distributed fiber optic sensing, the importance and challenges in railroad continuous CM, and the state-of-the-art application and future roadmap for the application of DAS in the railroad industry. [ABSTRACT FROM AUTHOR]

The use of Machine Learning (ML) has rapidly spread across several fields of applied sciences, having encountered many applications in Structural Dynamics and Vibroacoustic (SD&V). An advantage of ML algorithms compared to traditional techniques is that physical phenomena can be modeled using only sampled data from either measurements or simulations. This is particularly important in SD&V when the model of the studied phenomenon is either unknown or computationally expensive to simulate. This paper presents a survey on the application of ML algorithms in three classical problems of SD&V: structural health monitoring, active control of noise and vibration, and vibroacoustic product design. In structural health monitoring, ML is employed to extract damage-sensitive features from sampled data and to detect, localize, assess, and forecast failures in the structure. In active control of noise and vibration, ML techniques are used in the identification of state-space models of the controlled system, dimensionality reduction of existing models, and design of controllers. In vibroacoustic product design, ML algorithms can create surrogates that are faster to evaluate than physics-based models. The methodologies considered in this work are analyzed in terms of their strength and limitations for each of the three considered SD&V problems. Moreover, the paper considers the role of digital twins and physics-guided ML to overcome current challenges and lay the foundations for future research in the field. • Review of machine learning applied in vibroacoustic and structural dynamics. • Basic principles of the machine learning methods commonly used are presented. • Control, structural health monitoring and design are the most relevant applications. • The current implementation scenario and challenges are reviewed for each application. • This article sheds light on trends and research opportunities merging both fields. [ABSTRACT FROM AUTHOR]

Concerning the cost- and resource-saving maintenance of assembly products, it is vital to detect any potential malfunctions, defects or structural damage at the earliest-possible stage. For this reason, considerable efforts are being put into the development of Structural Health Monitoring, a field encompassing different approaches to damage identification and capable of preventing defects and even failure. Structural Health Monitoring is often supported by machine learning, a tool for rapid and effective damage identification that can recognize patterns or changes in the data received from the structure. Despite the advances machine learning has made in recent years, obtaining a suitable data set for the efficient training of machine learning algorithms within Structural Health Monitoring remains a challenge. Currently, the data are usually obtained experimentally, with numerical or analytical models. However, the experimental approach can often be time consuming, while the reliability of numerically obtained data relies heavily on the accuracy of the numerical models in capturing the true behavior of the structure. Analytical models may be constrained by the complexity of the observed object. In this paper an alternative approach based on an experimental–numerical (i.e., hybrid) modeling approach is proposed to build a training set for Structural Health Monitoring. Frequency Based Substructuring is utilized to determine the response model of the assembled system based on the properties of its components as well as to mix experimental and numerical models, while leveraging the advantages of each. This makes it possible to generate the samples of the training set in the form of hybrid models of the structure of interest, exhibiting the realistic properties of a physical structure, with a reasonable measurement effort. Here, the approach is demonstrated for the process of joint-damage identification. [Display omitted] • An approach to machine learning training set generation is proposed. • The proposed approach focuses on joint-damage identification. • Analytical, numerical and experimental data are used to obtain training data. • Dynamic substructuring is utilized to obtain training samples. • The proposed approach is tested in a case-study on a laboratory structure. [ABSTRACT FROM AUTHOR]

The vibration signal analysis-based structural health monitoring (SHM) system has been widely used for bridge condition identification. With the development of artificial intelligence (AI), machine learning (ML) has made numerous breakthroughs in the detection of civil engineering structures. However, conventional data-driven ML methods heavily rely on prior knowledge for their performance. This paper proposes a deep learning (DL) model 1DCNN-BiLSTM for detecting small local structural changes of reinforced concrete (RC) beams, which applies the structure of the Inception module in GoogLeNet to one-dimensional convolution neural networks (1DCNNs) for feature extraction at different scales, and combines the advantages of bidirectional long short-term memory (BiLSTM) modules for processing long time-series data. Firstly, a three-dimensional numerical model of the RC beam was established using finite element software, and the accuracy of the numerical model was verified by comparing it with the test results. Based on the numerical model, tests on RC beams impacted by falling hammers were carried out, and the acceleration signals of the beam in different states were collected as a dataset. The proposed DL model can automatically extract the spatial and temporal domain features in the signals, accurately identify the location of small bridge local variations, the accuracy in the test set is 98.8% compared to 92.6% for a traditional ML model, and has better noise immunity performance and robustness to the missing data than the traditional ML model. Finally, the internal inference process of the model was explored and visualized, illustrating that the model has adaptive learning capabilities. [ABSTRACT FROM AUTHOR]

• We present a multitask sparse Bayesian learning based damage localization and robust diagnostics method. • A guided-wave based multi-task model is constructed to improve the efficacy of damage localization. • A multi-task complex hierarchical sparse Bayesian learning model is presented, which allows the posterior uncertainty of damage localization to be effectively quantified. • The quantified posterior uncertainty of the model parameters is utilized to detect whether the inversion for damage localization is accurate. • Numerical and experimental studies are carried out to verify the effectiveness of the proposed method. The inversion of guided-wave data for accurate damage localization is a challenging problem when using guided waves for nondestructive testing and robust diagnostics. It is especially important to detect incorrect damage localization without knowing the original damage. In this paper, a new damage localization and robust diagnostics method is proposed. A Multi-task Complex Hierarchical Sparse Bayesian learning (MuCHSBL) algorithm is presented to solve the inverse problem for damage localization based on the data measured from a small number of sensors. The multi-task model improves the efficacy of damage localization by utilizing the consistency of damage locations for tasks with different signal frequencies. A Sparse Bayesian learning algorithm is also introduced to utilize the spatial sparsity of the damage, since structural damage typically occurs at only a few localized areas. The quantified posterior uncertainty of the model parameters gives a sense of confidence in the damage localization results. Utilizing the different levels of uncertainty in the optimal and suboptimal inversion models, diagnostic tools are proposed to detect whether the inversion for damage localization is accurate, without knowing the original damage. Numerical and experimental studies are carried out to verify the effectiveness of the proposed method. It is demonstrated that the damage localization efficacy of multi-task model is much higher than that of single-task model; moreover, the accuracy of damage localization results can be diagnosed effectively by the posterior uncertainty quantification of the model parameters. [ABSTRACT FROM AUTHOR]

Different health-monitoring techniques were considered in the literature to enhance the safety and stability of Connected Autonomous Vehicle (CAV) platoons. Mitigating these faults is faster and more reliable if the fault structure is known. In this paper, we propose using transmissibility operators, which are relationships that relate a set of velocities with another in the platoon, to classify the faults. Transmissibility operators were shown to be exceptional in signal estimation. However, it is also shown to be noncausal and thus can only be used offline. To this end, we propose using Data Aggregation (DAgger), an extension to imitation learning, to transfer the classification experience from transmissibility operators to a novice machine learning agent to be used online. The integration of transmissibility-DAgger gives the ability to adapt to new fault classes without the need to re-train the diagnosis model from the beginning. A heterogeneous CAV platoon was modeled with three different faults separately. These faults are actuator disturbances, false data injection attacks, and communication time delays. The classification scheme depends on estimating the faulty signal in the case of each fault class. Next, the measured faulty signal is compared with the three estimations, and the closer fault estimation to the measured one is considered the actual fault on the platoon. The proposed algorithm is then tested on the platoon model and then applied to an experimental setup that consists of three autonomous robots. The proposed results are compared with six different machine learning classification models. The overall classification accuracy achieved was 95.8% for the experiment using the proposed approach. [ABSTRACT FROM AUTHOR]

A primary motivation for the development and implementation of structural health monitoring systems, is the prospect of gaining the ability to make informed decisions regarding the operation and maintenance of structures and infrastructure. Unfortunately, descriptive labels for measured data corresponding to health-state information for the structure of interest are seldom available prior to the implementation of a monitoring system. This issue limits the applicability of the traditional supervised and unsupervised approaches to machine learning in the development of statistical classifiers for decision-supporting SHM systems. The current paper presents a risk-based formulation of active learning, in which the querying of class-label information is guided by the expected value of said information for each incipient data point. When applied to structural health monitoring, the querying of class labels can be mapped onto the inspection of a structure of interest in order to determine its health state. In the current paper, the risk-based active learning process is explained and visualised via a representative numerical example and subsequently applied to the Z24 Bridge benchmark. The results of the case studies indicate that a decision-maker's performance can be improved via the risk-based active learning of a statistical classifier, such that the decision process itself is taken into account. • A novel risk-based approach to active learning for SHM is proposed. • Statistical classifiers are developed with respect to decision-making. • Decision processes are modelled using influence diagrams. • Label querying is guided according to expected value of information. • Decision-making performance is shown to improve using this novel approach. [ABSTRACT FROM AUTHOR]

Predictive maintenance, as one of the core components of Industry 4.0, takes a proactive approach to maintain machines and systems in good order to keep downtime to a minimum and the airline maintenance industry is not an exception to this. To achieve this goal, practices in Structural Health Monitoring (SHM) complement the existing Non-Destructive-Testing (NDT) have been established in the last decades. Recently, the increasing computational capability such as utilization of a graphical processing unit (GPU) in combination with advanced machine learning techniques such as deep learning has been one of the main drivers in the advancement of predictive analytics in condition monitoring. In our previous work, we proposed a novel approach using deep learning for guided wave based structural health called DeepSHM. As a study case, we treated an ultrasonic signal from guided Lamb wave SHM with a convolutional neural network (CNN). In that work, we only considered a single central frequency excitation. This led to a single governing wavelength which is normally good for the detection of a single damage size. In classical signal processing, applying a broader excitation frequency poses an analysis and interpretation nightmare because it contains more complex information and thus is difficult to understand. This problem can be overcome with deep learning; however, it creates another problem: while deep learning typically results in a more accurate result prediction, it is specifically made for solving only certain types of tasks. While many papers have already introduced deep learning for diagnostics, many of these works are only proposing novel predictive techniques, however the mathematical formalization is lacking, and we are not informed about why we should treat acoustic signal with deep learning. So, the basis of 'explainable AI' for SHM and NDT is currently lacking. For this reason, in this paper, we would like to extend our previous work into a more generalized. Rather than focusing on a novel technique, we propose a plausible theoretical perspective inspired from neuroscience for signal representation of deep learning framework to model machine perception in structural health monitoring (SHM), especially because SHM typically involves multiple sensory input from different sensing locations. To do this, we created a set of artificial data from a finite element model (FEM) and represented DeepSHM in two different ways: 1). Perpetual representation of observation and 2). Hierarchical structure of entities that is decomposable in a smaller sub-entity. Consequently, we assume two plausible models for DeepSHM: 1). Either it behaves as a single deciding actor since the observation is regarded as perpetual, and 2). Or it acts as a multiple actor with independent outputs since multiple sensors can form different output probabilities. These artificial data were split into several different input representations, classified into several damage scenarios and then trained with commonly used deep learning training parameters. We compare the performance metrics of each perception model to describe the training behavior of both representations. [ABSTRACT FROM AUTHOR]

• This paper presents a deep learning network that inspects ultrasonic signals for defect detection. • The network is an adaptation of WaveNet, hence is based on causal dilated convolutional neural networks and residual blocks. • The network is shown to outperform both a widely used conventional analysis method and some competing deep learning algorithms. • The validation was performed on two datasets, one generated via finite element simulations on a steel plate, the other by actual experiments on a composite plate. • Significant improvements over the conventional method are obtained especially when testing occurs at temperatures well outside the range in the training set of signals. This paper presents a deep learning network that performs automatic detection of defects by inspecting full ultrasonic guided wave signals excited in plate structures. The findings show that the algorithm, which is an adaptation of WaveNet, and hence is based on causal dilated convolutional neural networks, is effectively able to learn features and/or patterns related to the presence of waves scattered from damage, thus eliminating the need for any feature engineering to be performed by human operators. The network outperformed the widely used conventional approach that combines the optimal baseline selection and baseline signal stretch compensation methods when tested on two different datasets. The first dataset consisted of finite element simulated Lamb wave signals acquired in a pitch-catch configuration on a steel plate across a 50 °C range of temperature variations, and the second was a publicly available experimental dataset of Lamb wave signals also acquired in pitch-catch mode on a composite plate with a 40 °C range of variations. The improvements over the conventional approach are particularly encouraging when analyzing signals at temperatures well outside the temperature range available in the set of baseline signals, hence suggesting that this class of algorithms can complement or substitute existing methods, especially when testing occurs at unseen environmental and operational conditions, or when the effects of sensor drift make conventional methods less effective. [ABSTRACT FROM AUTHOR]

• Contribution of attenuation mechanisms in propagation path can be assessed. • Utilising physics knowledge to inform learning benefits model generation. • Consideration of space over which physics is described can benefit model. The use of ultrasonic guided waves to probe the materials/structures for damage continues to increase in popularity for non-destructive evaluation (NDE) and structural health monitoring (SHM). The use of high-frequency waves such as these offers an advantage over low-frequency methods from their ability to detect damage on a smaller scale. However, in order to assess damage in a structure, and implement any NDE or SHM tool, knowledge of the behaviour of a guided wave throughout the material/structure is important (especially when designing sensor placement for SHM systems). Determining this behaviour is extremely difficult in complex materials, such as fibre–matrix composites, where unique phenomena such as continuous mode conversion takes place. This paper introduces a novel method for modelling the feature-space of guided waves in a composite material. This technique is based on a data-driven model, where prior physical knowledge can be used to create structured machine learning tools; where constraints are applied to provide said structure. The method shown makes use of Gaussian processes, a full Bayesian analysis tool, and in this paper it is shown how physical knowledge of the guided waves can be utilised in modelling using an ML tool. This paper shows that through careful consideration when applying machine learning techniques, more robust models can be generated which offer advantages such as extrapolation ability and physical interpretation. [ABSTRACT FROM AUTHOR]

The deployment of systems experiencing high-rate dynamic events, such as hypersonic vehicles, advanced weaponry, and active blast mitigation systems, require high-rate structural health monitoring (HRSHM) capabilities in the sub-millisecond realm to ensure continuous operations and safety. However, the development of high-rate feedback systems is a complex task because these dynamic systems are uniquely characterized by (1) large uncertainties in their external loads, (2) high levels of non-stationarity and heavy disturbance, and (3) unmodeled dynamics from changes in system configuration. In this paper, we present a deep learning algorithm specifically engineered for HRSHM applications. It consists of an ensemble of recurrent neural networks (RNNs) constructed with long short-term memory (LSTM) cells with transfer learning capabilities to cope with the highly limited availability of training data as it is typical for high-rate systems. The use of an ensemble of RNNs empowers multi-rate sampling capability to capture multi-temporal features of the time series, thus enabling modeling of non-stationarities. Also, because the RNNs use short-sequence LSTMs and are arranged in parallel, the computation time is substantially reduced to the sub-millisecond range. The performance of the algorithm is investigated on experimental high-rate dynamic data produced from accelerated drop tower tests and benchmarked against results from a prior investigation using a purely on-the-edge algorithm termed variable input space observer (VIO) and its hybrid architecture (hybrid VIO) that incorporated physical knowledge and is considered as an upper bound on the algorithm performance. Numerical results show that the ensemble of RNNs, composed of five RNNs, significantly outperformed the VIO, with performance close to that of the hybrid VIO when it comes to the estimation error metrics, with an average computation time of 25 μ s per prediction step. Yet, the ensemble of RNNs exhibits chattering for low-amplitude excitations, likely attributable to the behavior of RNNs sampling at small time delays. An examination of the RNN weights and hidden states confirms that the algorithm can capture multi-temporal features, and an investigation with respect to noise in the training dataset showed that the algorithm is robust for signal-to-noise ratios up to 20 dB. • Machine learning-based architecture for non-stationary time-series analysis. • Transfer learning for accelerated training with limited data. • Ensemble-based system estimation using parallel recurrent neural networks integrated with attention mechanism is proposed. • Sib-millisecond estimation of a high-rate system is achieved though the proposed architecture. [ABSTRACT FROM AUTHOR]

• Presented a probabilistic tensor learning approach for SHM data imputation and forecasting. • Proposed an efficient incremental learning scheme to deal with streaming data. • Developed a Bayesian tensor decomposition framework for missing data recovery. • Integrated tensor decomposition and vector autoregression for response forecasting. • Demonstrated effectiveness of the approach on a concrete bridge with over three-year records. There has been increased interest in missing sensor data imputation, which is ubiquitous in the field of structural health monitoring (SHM) due to discontinuous sensing caused by sensor malfunction. Recent development in Bayesian temporal factorization models for high-dimensional time series analysis has provided an effective tool to solve both imputation and prediction problems. However, for large datasets, the default Bayesian temporal factorization model becomes inefficient since the model has to be fully retrained when new data arrives. A potential solution is to train the model using a short time window covering only most recent data; however, by doing so, we may miss some critical dynamics and long-term dependencies which can only be identified from a longer time window. To address this fundamental issue in temporal factorization models, this paper presents an incremental Bayesian matrix/tensor learning scheme to achieve efficient imputation and prediction of structural response in long-term SHM. In particular, a spatiotemporal tensor is first constructed followed by Bayesian tensor factorization that extracts latent features for missing data imputation. To enable structural response forecasting based on long-term and incomplete sensing data, we develop an incremental learning scheme to effectively update the Bayesian temporal factorization model. The performance of the proposed approach is validated on continuous field-sensing data (including strain and temperature records) of a concrete bridge, based on the assumption that strain time histories are highly correlated to temperature recordings. The results indicate that the proposed probabilistic tensor learning framework is accurate and robust even in the presence of large rates of random missing, structured missing and their combination. The effect of rank selection on the imputation and prediction performance is also investigated. The results show that a better estimation accuracy can be achieved with a higher rank for random missing whereas a lower rank for structured missing. [ABSTRACT FROM AUTHOR]