Unsupervised Data Anomaly Detection Based on Spatio-Temporal Sparse Integration Framework

Guided wave-based structural health monitoring is extensively utilized in various industrial applications to ensure the integrity of components within industrial systems. Among these monitoring techniques, principal component analysis (PCA) reconstruction methods are widely used for anomaly detection due to their computational efficiency and interoperability. However, existing PCA reconstruction methods are semisupervised anomaly detection approaches that require training on historical normal data and fail to detect anomalous signals within the training set. To address this limitation, this work proposes a correntropy PCA (C-PCA), enabling fully unsupervised anomaly detection on raw training data without requiring label information, when the dataset contains a high proportion of abnormal signals. This method allows anomaly detection on real-time measurements without the need for precleaned historical normal data or can also be used to generate clean data for existing semisupervised anomaly detection methods. In correntropy PCA, principal components are extracted from the correntropy matrix rather than the correlation matrix. The correntropy, representing the statistical dependence between samples of guided waves, is estimated utilizing a Gaussian kernel with a specified kernel width. Through the optimization of the kernel width, the correntropy PCA reconstruction method demonstrates superior anomaly detection performance compared with the standard PCA reconstruction method, especially in scenarios where training data are contaminated by a significant proportion of abnormal signals. Guidelines for the optimization of the kernel width are provided. The effectiveness of the correntropy PCA reconstruction-based anomaly detection method is validated using data collected from ten regions over an 80-day period, encompassing guided waves induced by damage occurring over durations ranging from 2 to 20 days.

Modal tracking plays a vital role in structural health monitoring since changes in modal parameters help us understand a structure’s dynamic characteristics and may reflect the potential deterioration of structural performance. Although numerous modal parameter estimation (MPE) methods exist, it is not guaranteed that an MPE process will exclude all spurious modes and not lose any physical modes every time over a long-term monitoring period. Relatively large damping of a structure, poor data quality, and significant changes in structural modal parameters may make the estimated modal parameters spurious, missing, or misclassified. It makes long-term modal tracking semiautomated or manual, which constrains timely downstream applications such as anomaly detection, condition assessment, and decision making. This research aims to propose a long-term continuous automatic modal tracking algorithm based on Bayesian inference even when the modal parameters, damping, and data quality change significantly. Bayesian inference is used to determine the physical modes from the results of existing MPE methods. Both the modes identified from the most recent response set and the modal probability model from multiple previous response sets are considered in the Bayesian model to better determine the physical modes from the results of MPE. Moreover, the proposed algorithm requires only three extra hyperparameters compared to general modal tracking algorithms, and they can be quickly determined by a grid search method. The performance of the proposed algorithm is verified by a numerical example and a real-world civil structure Z24 Bridge benchmark.

Anomaly detection is the process of identifying unexpected items or events in datasets, which differ from the norm. In contrast to standard classification tasks, anomaly detection is often applied on unlabeled data, taking only the internal structure of the dataset into account. This challenge is known as unsupervised anomaly detection and is addressed in many practical applications, for example in network intrusion detection, fraud detection as well as in the life science and medical domain. Dozens of algorithms have been proposed in this area, but unfortunately the research community still lacks a comparative universal evaluation as well as common publicly available datasets. These shortcomings are addressed in this study, where 19 different unsupervised anomaly detection algorithms are evaluated on 10 different datasets from multiple application domains. By publishing the source code and the datasets, this paper aims to be a new well-funded basis for unsupervised anomaly detection research. Additionally, this evaluation reveals the strengths and weaknesses of the different approaches for the first time. Besides the anomaly detection performance, computational effort, the impact of parameter settings as well as the global/local anomaly detection behavior is outlined. As a conclusion, we give an advise on algorithm selection for typical real-world tasks.

Supervised deep learning approaches provide state-of-the-art performance on medical image classification tasks for disease screening. However, these methods require large labeled datasets that involve resource-intensive expert annotation. Further, disease screening applications have low prevalence of abnormal samples; this class imbalance makes the task more akin to anomaly detection. While the machine learning community has proposed unsupervised deep learning methods for anomaly detection, they have yet to be characterized on medical images where normal vs. anomaly distinctions may be more subtle and variable. In this work, we characterize existing unsupervised anomaly detection methods on retinal fundus images, and find that they require significant fine tuning and offer unsatisfactory performance. We thus propose an efficient and effective transfer-learning based approach for unsupervised anomaly detection. Our method employs a deep convolutional neural network trained on ImageNet as a feature extractor, and subsequently feeds the learned feature representations into an existing unsupervised anomaly detection method. We show that our approach significantly outperforms baselines on two natural image datasets and two retinal fundus image datasets, all with minimal fine-tuning. We further show the ability to leverage very small numbers of labelled anomalies to improve performance. Our work establishes a strong unsupervised baseline for image-based anomaly detection, alongside a flexible and scalable approach for screening applications.

Recently developed piezoceramic-based transducers, known as smart aggregates (SAs), have shown their applicability and versatility in various applications of structural health monitoring (SHM). The lead zirconate titanate (PZT) patches embedded inside SAs have different modes that are more suitable for generating or receiving different types of stress waves (e.g. P and S waves, each of which has a unique role in SHM). However, due to the geometry of the 2D PZT patch, the embedded SA can only generate or receive the stress wave in a single direction and thus greatly limits its applications. This paper is the first of a series of two companion papers that introduces the authors’ latest work in developing a novel, embeddable spherical smart aggregate (SSA) for the health monitoring of concrete structures. In addition to the 1D guided wave produced by SA, the SSA embedded in concrete structures can generate or receive omni-directional stress waves that can significantly improve the detection aperture and provide additional functionalities in SHM. In the first paper (Part I), the detailed fabrication procedures with the help of 3D printing technology and electrical characterization of the proposed SSA is presented. The natural frequencies of the SSA were experimentally obtained and further compared with the numerical results. In addition, the influence of the components’ thickness (spherical piezoceramic shell and epoxy) and outer radius (spherical piezoceramic shell and protection concrete) on the natural frequencies of the SSA were analytically studied. The results will help elucidate the key parameters that determine the natural frequencies of the SSA. The natural frequencies of the SSA can thus be designed for suitability in the damage detection of concrete structures. In the second paper (Part II), further numerical and experimental verifications on the performance of the proposed SSA in concrete structures will be discussed.

Unsupervised anomaly detection (UAD) attracts a lot of research interest and drives widespread applications, where only anomaly-free samples are available for training. Some UAD applications intend to locate the anomalous regions further even without any anomaly information. Although the absence of anomalous samples and annotations deteriorates the UAD performance, an inconspicuous, yet powerful statistics model, the normalizing flows, is appropriate for anomaly detection (AD) and localization in an unsupervised fashion. The flow-based probabilistic models, only trained on anomaly-free data, can efficiently distinguish unpredictable anomalies by assigning them much lower likelihoods than normal data. Nevertheless, the size variation of unpredictable anomalies introduces another inconvenience to the flow-based methods for high-precision AD and localization. To generalize the anomaly size variation, we propose a novel multiscale flow-based framework (MSFlow) composed of asymmetrical parallel flows followed by a fusion flow to exchange multiscale perceptions. Moreover, different multiscale aggregation strategies are adopted for image-wise AD and pixel-wise anomaly localization according to the discrepancy between them. The proposed MSFlow is evaluated on three AD datasets, significantly outperforming existing methods. Notably, on the challenging MVTec AD benchmark, our MSFlow achieves a new state-of-the-art (SOTA) with a detection AUORC score of up to 99.7%, localization AUCROC score of 98.8% and PRO score of 97.1%.

Damage imaging in structural health monitoring with fine-tuned conditional diffusion model

Acoustic source localization using L-shaped sensor clusters: A review

Although the interest in the Digital Twin (DT) has grown across industry and academia, only a few concrete proposals for the technical implementation of such a DT have been made. To benefit from the operational assessments and improvements through the use of a DT, the integration of structural analysis should be addressed. To connect DTs to their physical counterparts during the whole service life, changes during the lifetime, e.g., due to fatigue or damage, must be detected, quantified and evaluated. The suitable tool to achieve this goal is Structural Health Monitoring (SHM). This contribution examines the state of the art for DTs in the context of structural applications. Furthermore, the role of SHM as a necessary condition for the realisation of DTs is discussed. With the help of the SHM levels according to Rytter, a method is proposed to select the required information depth for adapting the DT to a changing physical twin. Only after detecting a measurable deviation (Level 1) it is necessary to apply more detailed models to assess changes in the structural behaviour. These considerations are illustrated by the example of a cantilever beam structure.KeywordsDigital TwinStructural Health MonitoringIntelligent structuresStructural damage indicatorsModelling

Effectively monitoring the operational status of reactor coolant pumps (RCPs) is crucial for enhancing the safety and stability of nuclear power operations. To address the challenges of limited interpretability and suboptimal detection performance in existing methods for detecting abnormal operating states of RCPs, this paper proposes an interpretable, unsupervised anomaly detection approach. This innovative method designs a framework that combines Kernel Self-Organizing Map (Kernel SOM) clustering with Bayesian Posterior Inference. Specifically, the proposed method uses Kernel SOM to extract typical patterns from normal operation data. Subsequently, a distance probability distribution model reflecting the data distribution structure within each cluster is constructed, providing a robust tool for data distribution analysis for anomaly detection. Finally, based on prior knowledge, such as distance probability distribution, the Bayesian Posterior Inference is employed to infer the probability of the equipment being in a normal state. By constructing distribution models that reflect data distribution structures and combining them with posterior inference, this approach realizes the traceability and interpretability of the anomaly detection process, improving the transparency of anomaly detection and enabling operators to understand the decision logic and the analysis of the causes of anomalous occurrences. Verification via real-world operational data demonstrates the method’s superior effectiveness. This work offers a highly interpretable solution for RCP anomaly detection, with significant implications for safety-critical applications in the nuclear energy sector.

To defend a network system from security risks, intrusion detection systems (IDSs) have been playing an important role in recent years. There are two types of detection algorithms of IDSs: misuse detection and anomaly detection. Because misuse detection is based on a signature which is created from the features of attack traffic by security experts, it can achieve accurate and stable detection. However, its weakness is the difficulty of detecting new attacks (i.e., 0-day attack), and the cost of maintaining the latest signature version. Thinking of the increase of the skillful intrusion, e.g., intrusion showing similar access behavior to normal, misuse detection cannot handle these critical attacks, which results in a large number of false alarms. To cope with these problems, we present a clustering algorithm based on an unsupervised anomaly detection. We evaluated our system using Kyoto2006+ data set and KDD Cup 1999 data set. Evaluation results show that our approach achieved a higher detection rate in the region of very low false positive rate and real-time preprocessing capability.

BackgroundChemical exchange saturation transfer (CEST) magnetic resonance imaging (MRI) demonstrates significant potential for early disease detection, yet encounters challenges in lesion analysis due to interference from concomitant effects during imaging. Although machine learning-based unsupervised anomaly detection (UAD) methods offer feasible solutions for identifying subtle lesions in contaminated images, current UAD approaches primarily dependent on convolutional neural network (CNN) or Transformer architectures to process image-level data from conventional imaging modalities, exhibit limitations when applied to CEST data characterized by spectral information. This study aimed to propose a novel UAD framework tailored for CEST spectral data to overcome these limitations, enabling precise pixel-wise anomaly detection for early-stage lesions.MethodsThe proposed framework named Mamba-AE employs a multi-layers encoder-decoder architecture with stacked Mamba blocks as the core component. Each Mamba block integrates three key modules: selective state-space models (SSMs) to dynamically adjust parameters based on input sequences, enabling adaptive long-range dependency modeling of CEST spectral data; Gated Multi-Layer Perceptron with Sigmoid Linear Unit (SiLU) activation and residual connections to enhance non-linear feature learning and stability; Multi-Scale Feature Alignment that aligns hierarchical encoder-decoder features via cosine similarity, preserving physiological semantics across scales. The framework adopts a dual-domain reconstruction strategy: data-space reconstruction minimizes pixel-wise spectral errors through Huber loss, while feature-space reconstruction enforces consistency between paired encoder-decoder layer features via cosine similarity loss. Anomaly scores are generated by combining normalized residuals from data-space errors and discrepancies in feature-space alignment, enabling precise pixel-level lesion localization in early-stage CEST data. For quantitative evaluation of lesion detection performance, the area under the curve (AUC) and Dice similarity coefficient are employed as core metrics. The framework was trained exclusively on CEST spectra from healthy rat brain tissue and validated on an independent dataset from a rat model of transient focal cerebral ischemia induced by middle cerebral artery occlusion (MCAO) with reperfusion.ResultsThe proposed method has been validated on ischemic stroke rat datasets at 2, 6, and 24 hours post-occlusion. Mamba-AE demonstrated superior performance across all stages. At the critical 2-hour time point, it achieved the highest performance among all methods, with an AUC of 95.91% and a Dice score of 88.43% (sensitivity >89%, specificity >92%). The advantage remained substantial at 6 hours (AUC: 92.19%, Dice: 85.23%) and 24 hours (AUC: 94.64%, Dice: 87.79%), consistently outperforming other approaches. Furthermore, anomaly heatmaps revealed precise lesion localization capabilities, with spatial accuracy correlating strongly with histopathological ground truth measurements.ConclusionsMamba-AE provides a computationally efficient and robust solution for early disease detection in CEST MRI. Its integration of spectral feature learning and structural preservation highlights its potential for clinical applications requiring high sensitivity to subtle pathological changes.

Vibration-based damage detection methods, leveraging structural response data to identify structural damage, play an important role in structural health monitoring, particularly for critical infrastructure like bridges and high-rise buildings. In recent years, deep learning, which can autonomously learn damage-sensitive features and classify damage from complex datasets, has been widely used in structural damage detection. This paper reviews deep learning architectures, including Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and autoencoders, applied in vibration-based structural damage detection for civil structures over the past ten years, highlighting their advantages and disadvantages. The challenges and future trends for vibration-based damage detection by using deep learning are summarized.

Application of massive data to structural health monitoring (SHM) may lead to serious problems such as difficulty, computational inefficiency, and low damage detectability. This paper proposes innovative hybrid methods in order to detect damage under massive data, ambient vibration, and environmental and/or operational variability. Each of the proposed methods consists of a three-stage algorithm including response modelling via a time series representation, Gaussian mixture model (GMM) for dimensionality reduction, and an outlier detector. Due to the importance of response modelling under ambient vibration, this article employs the combination of AR with ARX called the ARARX model. In the second stage, a GMM is individually fitted to residual samples extracted from ARARX regarding the undamaged and damaged conditions to obtain low-dimensional features from massive data. Eventually, outlier analyses are carried out by the Mahalanobis distance and an auto-associative neural network to make a decision about the occurrence of damage. Two different kinds of threshold limits are also considered to examine the performances of the proposed methods. A large-scale cable-stayed bridge is applied to validate the reliability and effectiveness of these methods with several comparative studies. Results demonstrate that both proposed methods are effective tools for damage detection.

Anomaly detection (AD) aims to distinguish the data points that are inconsistent with the overall pattern of the data. Recently, unsupervised anomaly detection methods have aroused huge attention. Among these methods, feature representation (FR) plays an important role, which can directly affect the performance of anomaly detection. Sparse representation (SR) can be regarded as one of matrix factorization (MF) methods, which is a powerful tool for FR. However, there are some limitations in the original SR. On the one hand, it just learns the shallow feature representations, which leads to the poor performance for anomaly detection. On the other hand, the local geometry structure information of data is ignored. To address these shortcomings, a graph regularized deep sparse representation (GRDSR) approach is proposed for unsupervised anomaly detection in this work. In GRDSR, a deep representation framework is first designed by extending the single layer MF to a multilayer MF for extracting hierarchical structure from the original data. Next, a graph regularization term is introduced to capture the intrinsic local geometric structure information of the original data during the process of FR, making the deep features preserve the neighborhood relationship well. Then, a L1-norm-based sparsity constraint is added to enhance the discriminant ability of the deep features. Finally, a reconstruction error is applied to distinguish anomalies. In order to demonstrate the effectiveness of the proposed approach, we conduct extensive experiments on ten datasets. Compared with the state-of-the-art methods, the proposed approach can achieve the best performance.