Search Results (689)

Against the backdrop of global climate change, remote subtropical mountainous power grids face severe operational challenges due to their fragile infrastructure and complex climatic conditions. However, existing research has insufficiently addressed load forecasting in data-sparse regions, particularly lacking systematic analysis of the “meteorology–load–failure” coupling mechanism. To address this gap, this study focused on 10 kV distribution lines in a typical subtropical monsoon region of southern China. Based on hourly load and meteorological data from 2016 to 2025, we propose a two-stage hybrid model combining “Random Forest (RF) feature selection + Long Short-Term Memory (LSTM) time series forecasting”. Through deep feature engineering, composite, lagged, and interactive features were constructed. Using the RF algorithm, we quantitatively identified the core drivers of load variation across different time scales: at the hourly scale, variations are dominated by historical inertia (with weights of 0.5915 and 0.3757 for 1-h and 24-h lagged loads, respectively); at the daily scale, the logic shifts to meteorological triggering and cumulative effects, where the composite feature load_lag1_hi_product emerged as the most critical driver (weight of 0.8044). Experimental results demonstrate that the hybrid model significantly improved forecasting accuracy compared to the full-feature LSTM benchmark: on a daily scale, RMSE decreased by 13.29% and MAE by 16.67%, with R² reaching 0.8654; on an hourly scale, R² reached 0.9687. Furthermore, correlation analysis with failure data revealed that most grid faults occurred during intervals of extremely low load variation (0–5%), suggesting that “chronic stress” from environmental exposure in hot and humid conditions is the primary cause, with lightning identified as the leading external threat (26.90%). The interpretable forecasting framework proposed in this study transcends regional limitations. It provides a strategic “low-cost, high-resilience” prototype applicable to power systems in humid-subtropical zones worldwide, particularly for developing regions facing the dual challenges of data sparsity and climate vulnerability. Full article

Unplanned downtime in industrial machinery remains a major challenge, causing substantial economic losses and safety risks across sectors such as manufacturing, food processing, oil and gas, and transportation. This systematic review investigates the application of machine learning (ML) techniques for anomaly and fault detection within the broader context of predictive maintenance. Following a hybrid review methodology, relevant studies published between 2010 and 2025 were collected from major databases including IEEE Xplore, ScienceDirect, SpringerLink, Scopus, Web of Science, and arXiv. The review categorizes approaches into supervised, unsupervised, and hybrid paradigms, analyzing their pipelines from data collection and preprocessing to model deployment. Findings highlight the effectiveness of deep learning architectures such as convolutional neural networks (CNNs), recurrent neural networks (RNNs), autoencoders, and hybrid frameworks in detecting faults from time series and multimodal sensor data. At the same time, key limitations persist, including data scarcity, class imbalance, limited generalizability across equipment types, and a lack of interpretability in deep models. This review concludes that while ML-based predictive maintenance systems are enabling a transition from reactive to proactive strategies, future progress requires improved hybrid architectures, Explainable AI, and scalable real-time deployment. Full article

Timely identification of series arc faults (SAFs) is of vital importance for preventing electrical fires. How to identify SAFs at the output side of a frequency converter (i.e., output-side SAF) is still not clear. A new approach of identifying output-side SAFs by analyzing the output current signals from frequency converters was proposed. First, output-side SAF experiments were performed with harmonic power supplies. Second, the output current signals were decomposed into eight modal components by empirical wavelet transform and an autoregressive model was established. The autoregressive model parameters and the energy ratios of the first three modal components were adopted as the fault features. Finally, an optimized support vector machine was designed and employed to identify SAFs. Comparison tests with similar methods were performed and performance tests under different noise levels and operation conditions were conducted. The test results indicated that the proposed scheme can effectively recognize output-side SAFs. Its runtime is shorter than 1.4 ms. This method provides a reference for the development of industrial three-phase SAF detection devices. Full article

The deployment of 5G (Fifth-Generation) networks in industrial Internet of Things (IoT), intelligent transportation, and emergency communications introduces heterogeneous and dynamic network states, leading to frequent and diverse faults. Traditional fault detection methods typically emphasize either local temporal anomalies or global distributional characteristics, but rarely achieve an effective balance between the two. In this paper, we propose a parallel multi-branch convolutional neural network (CNN)–Transformer framework (MBCT) to improve fault diagnosis accuracy in 5G networks. Specifically, MBCT takes time-series network key performance indicator (KPI) data as input for training and performs feature extraction through three parallel branches: a CNN branch for local patterns and short-term fluctuations, a Transformer encoder branch for cross-layer and long-term dependencies, and a statistical branch for global features describing quality-of-experience (QoE) metrics. A gating mechanism and feature-weighted fusion are applied outside the branches to adjust inter-branch weights and intra-branch feature sensitivity. The fused representation is then nonlinearly mapped and fed into a classifier to generate the fault category. This paper evaluates the performance of the proposed model on both the publicly available TelecomTS multi-modal 5G network observability dataset and a self-collected SDR5GFD dataset based on software-defined radio (SDR). Experimental results demonstrate that the proposed model achieves superior performance in fault classification, achieving 87.7% accuracy on the TelecomTS dataset and 86.3% on the SDR5GFD dataset, outperforming the baseline models CNN, Transformer, and Random Forest. Moreover, the model contains approximately 0.57M parameters and requires about 0.3 MFLOPs per sample for inference, making it suitable for large-scale online fault diagnosis. Full article

In autonomous manufacturing systems, the performance of time-series-based anomaly detection and fault diagnosis is highly sensitive to window size selection. Conventional approaches rely on empirical rules or fixed window settings, which often fail to capture the diverse temporal characteristics of anomalies and lead to performance degradation. This study systematically addresses the window size selection problem by categorizing anomaly patterns into three representative types: variability, cycle, and local spike. Each pattern is associated with a distinct temporal scale and underlying physical mechanism. Based on this insight, an Anomaly Deviation-Based Window Size Selection (ADW) method is proposed, which quantitatively evaluates anomaly deviation as a function of window size. Unlike traditional preprocessing-oriented approaches, the proposed method redefines window size as a core design variable that directly governs anomaly representation and diagnostic sensitivity. The effectiveness of the ADW method is validated using tension data from a roll-to-roll continuous manufacturing process and vibration data from a rotating bearing fault dataset. Experimental results demonstrate that the proposed approach consistently identifies optimized window sizes tailored to different anomaly types, leading to improved fault classification accuracy and diagnostic robustness. The proposed framework provides a physically interpretable and data-driven guideline for adaptive window size selection in long-term autonomous manufacturing systems. Full article

The Linfen–Yuncheng Basin is located on the southern edge of the Fenwei Fault Zone, influenced by intense tectonic activity, thick Quaternary sedimentation, and anthropogenic disturbance, it exhibits prominent characteristics of ground subsidence and fissure development. However, uncertainties still exist regarding the primary controlling factors of subsidence. This study employs multi-temporal InSAR data, combined with small baseline subset (SBAS–InSAR) technology to invert the high-precision ground line of sight deformation fields, and conducts time-series decomposition analysis using the Seasonal Trend Decomposition (STL) method. The results show that from 2017 to 2025, subsidence was mainly concentrated in the central and southern regions of the basin, with a maximum cumulative subsidence exceeding 200 mm and an average annual subsidence rate of −40 mm/year. Its spatial distribution is highly consistent with major structural zones such as the Zhongtiao Mountain Front Fault and the Linyi Fault, indicating that fault activity exerts a significant controlling effect on subsidence patterns. Groundwater level fluctuations are positively correlated with overall ground subsidence, and the response rate of different monitoring points is constrained by differences in aquifer depth and permeability. Groundwater aquifer points exhibit rapid and reversible subsidence response, while confined aquifer points are affected by low-permeability or compressible layers, showing a significant lag effect. The research results indicate that time-series analysis based on InSAR can not only effectively reveal the subsidence evolution process at different scales, but also provide a scientific basis for groundwater resource regulation, geological disaster prevention and control, and sustainable regional land utilization. Full article

The Sındırgı earthquake sequence, with moment magnitudes of 6.1 on 10 August and 27 October 2025, respectively, occurred within the Simav Fault Zone in western Türkiye, rupturing nearby but structurally distinct fault segments. In this study, we combine Sentinel-1 InSAR time-series measurements with seismological data, geomorphic observations, and post-event field surveys to examine how deformation evolved between and after these events. InSAR results indicate coseismic line-of-sight displacements of 6–7 cm, followed by post-seismic deformation that persisted for months at 8–10 mm/yr. This behavior signifies that deformation continued well beyond the initial rupture. The estimated displacement does not align with a single fault plane. Instead, it corresponds to a network of early-mapped and previously unrecognized fault segments. Seismicity patterns and stress tensor inversions show that activity migrated spatially after 10 August and that the faulting mechanism altered before the second earthquake. When synthesized, observations indicate stress transfer within a modular, segmented fault system, thought to have been influenced by regional structural complexity. Field investigations after the October earthquake reported new surface cracks and fault traces, providing evidence of shallow deformation. The collected results indicate that post-seismic stress redistribution played a leading role in modulating the 2025 Sındırgı earthquake sequence. Full article

Modern industrial equipment is a cyber-physical system (CPS) consisting of physical production components and digital controls. Lowering maintenance costs and increasing availability is important to improve its efficiency. Modern methods, based on solving event prediction problem, in particular, prediction of remaining useful life (RUL), are used as a crucial step in a framework of reliability-centered maintenance to increase efficiency. But modern methods of RUL forecasting fall short when dealing with real-world scenarios, where CPS are described by multidimensional continuous high-frequency data with working cycles with variable duration. To overcome this problem, we propose a new method for fault prediction, which is based on extraction of semantic state vectors (SSVs) from working cycles of equipment. To implement SSV extraction, a new method, based on convolutional autoencoder and extraction of hidden state, is proposed. In this method, working cycles are detected in input data stream, and then they are converted to images, on which an autoencoder is trained. The output of an intermediate layer of an autoencoder is extracted and processed into SSVs. SSVs are then combined into a time series on which RUL is forecasted. After optimization of hyperparameters, the proposed method shows the following results: RMSE = 1.799, MAE = 1.374. These values are significantly more accurate than those obtained using existing methods: RMSE = 14.02 and MAE = 10.71. Therefore, SSV extraction is a viable technique for forecasting RUL. Full article

Industrial aluminum-block heating processes exhibit nonlinear dynamics, substantial time delays, and stringent requirements for fault detection and diagnosis, especially in semiconductor manufacturing and other high-precision electronic processes, where slight temperature deviations can accelerate device degradation or even cause catastrophic failures. To address these challenges, this study presents a digital twin-based intelligent heating platform for aluminum blocks with a dual-artificial-intelligence framework (dual-AI) for control and diagnosis, which is applicable to multi-port aluminum-block heating systems. The system enables real-time observation and simulation of high-temperature operational conditions via virtual-real interaction. The platform precisely regulates a nonlinear temperature control system with a prolonged time delay by integrating a conventional proportional–integral–derivative (PID) controller with a Levenberg–Marquardt-optimized backpropagation (LM-optimized BP) neural network. Simultaneously, a relay is employed to sever the connection to the heater, thereby simulating an open-circuit fault. Throughout this procedure, sensor data are gathered simultaneously, facilitating the creation of a spatiotemporal time-series dataset under both normal and fault conditions. A one-dimensional convolutional neural network (1D-CNN) is trained to attain high-accuracy fault detection and localization. PID+LM-BP achieves a response time of about 200 s in simulation. In the 100 °C to 105 °C step experiment, it reaches a settling time of 6 min with a 3 °C overshoot. Fault detection uses a 0.38 °C threshold defined based on the absolute minute-to-minute change of the 1-min mean temperature. Full article

On 7 January 2025, an Mw 7.0 earthquake struck Dingri County, Shigatse, Tibet. This was the largest event in the region in recent years. Analysis of the Dingri earthquake is urgent for understanding the coseismic slip and early postseismic deformation characteristics. In this study, the coseismic characteristics were analyzed by using Lutan-1 and Sentinel-1 data with the Differential Interferometric Synthetic Aperture Radar method, and then the Okada elastic half-space dislocation model was used to invert the coseismic slip distribution of the seismogenic fault. The postseismic characteristics were analyzed by Sentinel-1 ascending and descending orbits, then time-series deformation results were obtained with the Small Baseline Subset InSAR method. The main results are as follows: (1) The maximum coseismic subsidence is −2.03 m and the maximum coseismic uplift is 0.68 m, the coseismic deformation is concentrated on the west side of the new rupture trace generated by the coseismic events; (2) the ruptured fault is dominated by normal faulting with a minor strike-slip component, and the slip is mainly distributed at depths of 0–15 km, with a maximum slip of about 3.97 m; (3) the deformation characteristics of the fault in the postseismic stage are basically consistent with those during the coseismic stage. The research results play an important role in understanding the earthquake fault tectonic activities. Full article

Surface deformation induced by the extraction of natural resources constitutes a non-stationary spatiotemporal process. Modeling surface deformation time series obtained through Interferometric Synthetic Aperture Radar (InSAR) technology using deep learning methods is crucial for disaster prevention and mitigation. However, the complexity of model hyperparameter configuration and the lack of interpretability in the resulting predictions constrain its engineering applications. To enhance the reliability of model outputs and their decision-making value for engineering applications, this study presents a workflow that combines a Tree-structured Parzen Estimator (TPE)-based Bayesian optimization approach with ensemble inference. Using the Rhineland coalfield in Germany as a case study, we systematically evaluated six deep learning architectures in conjunction with various spatiotemporal coding strategies. Pairwise comparisons were conducted using a Welch t-test to evaluate the performance differences across each architecture under two parameter-tuning approaches. The Benjamini–Hochberg method was applied to control the false discovery rate (FDR) at 0.05 for multiple comparisons. The results indicate that TPE-optimized models demonstrate significantly improved performance compared to their manually tuned counterparts, with the ResNet+Transformer architecture yielding the most favorable outcomes. A comprehensive analysis of the spatial residuals further revealed that TPE optimization not only enhances average accuracy, but also mitigates the model’s prediction bias in fault zones and mineralize areas by improving the spatial distribution structure of errors. Based on this optimal architecture, we combined the ten highest-performing models from the optimization stage to generate a quantile-based susceptibility map, using the ensemble median as the central predictor. Uncertainty was quantified from three complementary perspectives: ensemble spread, class ambiguity, and classification confidence. Our analysis revealed spatial collinearity between physical uncertainty and absolute residuals. This suggests that uncertainty is more closely related to the physical complexity of geological discontinuities and human-disturbed zones, rather than statistical noise. In the analysis of super-threshold probability, the threshold sensitivity exhibited by the mining area reflects the widespread yet moderate impact of mining activities. By contrast, the fault zone continues to exhibit distinct high-probability zones, even under extreme thresholds. It suggests that fault-controlled deformation is more physically intense and poses a greater risk of disaster than mining activities. Finally, we propose an engineering decision strategy that combines uncertainty and residual spatial patterns. This approach transforms statistical diagnostics into actionable, tiered control measures, thereby increasing the practical value of susceptibility mapping in the planning of natural resource extraction. Full article

With the rapid evolution of cloud-native platforms, microservice-based systems have become increasingly large-scale and complex, making fast and accurate root cause localization and recovery a critical challenge. Runtime signals in such systems are inherently multimodal—combining metrics, logs, and traces—and are intertwined through deep, dynamic service dependencies, which often leads to noisy alerts, ambiguous fault propagation paths, and brittle, manually curated recovery playbooks. To address these issues, we propose GALR, a graph- and LLM-based framework for root cause localization and recovery in microservice-based business middle platforms. GALR first constructs a multimodal service call graph by fusing time-series metrics, structured logs, and trace-derived topology, and employs a GAT-based root cause analysis module with temporal-aware edge attention to model failure propagation. On top of this, an LLM-based node enhancement mechanism infers anomaly, normal, and uncertainty scores from log contexts and injects them into node representations and attention bias terms, improving robustness under noisy or incomplete signals. Finally, GALR integrates a retrieval-augmented LLM agent that retrieves similar historical cases and generates executable recovery strategies, with consistency checking against expert-standard playbooks to ensure safety and reproducibility. Extensive experiments on three representative microservice datasets demonstrate that GALR consistently achieves superior Top-k accuracy and mean reciprocal rank for root cause localization, while the retrieval-augmented agent yields substantially more accurate and actionable recovery plans compared with graph-only and LLM-only baselines, providing a practical closed-loop solution from anomaly perception to recovery execution. Full article

In the field of sensors, time series classification holds significant importance for applications such as industrial monitoring, mechanical fault diagnosis, and action recognition. However, while existing models demonstrate excellent classification accuracy, they generally suffer from insufficient interpretability. Shapelet-based methods offer interpretability advantages, yet existing models like ShapeFormer suffer from high computational resource consumption and low training efficiency during shapelet discovery and training phases, limiting their applicability in complex sensor time series classification tasks. To address this, our research proposes Efficiency ShapeFormer (EffiShapeFormer), an efficient time series classification framework, based on the latest shapelet model ShapeFormer. During the Shapelet Discovery phase, EffiShapeFormer introduces a dual-filtering mechanism. The Coarse Screening module efficiently identifies discriminative shapelets, while the Class-specific Representation module models these features to extract class-specific characteristics. Subsequently, in the Generic Representation stage, the proposed Convolution-Inverted Attention (CIA) module achieves synergistic integration of local feature extraction and global dependency modeling to capture cross-category generic features. Finally, the model fuses class-specific and generic features to achieve efficient and accurate time series classification. Experimental results on 22 sensor time series datasets demonstrate that EffiShapeFormer achieves higher average accuracy and F1-scores than baseline models, validating the proposed method’s significant advantages in both efficiency and performance. Full article

Predictive maintenance (PdM) often fails to progress beyond pilot projects because machine learning-based anomaly detection requires expert knowledge, extensive tuning, and labeled fault data. This paper presents an automated prototype that builds and evaluates multiple anomaly detection models with minimal manual configuration. The prototype automates feature creation, model training, hyperparameter search, and ensemble construction, while allowing domain experts to control how anomaly alerts are triggered and how detected events are reviewed. Developed in a multi-year photovoltaic (PV) solar farm case study, it targets operational anomalies such as sudden drops, underperformance periods, and abnormal drifts, using expert validation and synthetic benchmarks to shape and evaluate anomaly categories. Experiments on the real PV data, a synthetic PV benchmark, and a machine temperature dataset from the Numenta Anomaly Benchmark show that no single model performs best across datasets. Instead, diverse base models and both rule-based and stacked ensembles enable robust configurations tailored to different balances between missed faults and false alarms. Overall, the prototype offers a practical and accessible path toward PdM adoption by lowering technical barriers and providing a flexible anomaly detection approach that can be retrained and transferred across industrial time-series datasets. Full article

Drivetrain systems operate under varying operating conditions (OCs), which often obscure early-stage fault signatures and hinder robust condition monitoring (CM). This work introduces an AI digital platform developed during the EEDRIVEN project, featuring a holistic CM framework that integrates statistical time series methods—using Generalized AutoRegressive (GAR) models in a multiple model fault diagnosis scheme—with deep learning approaches, including autoencoders and convolutional neural networks, enhanced through a dedicated decision fusion methodology. The platform addresses all key CM tasks, including fault detection, fault type identification, fault severity characterization, and remaining useful life (RUL) estimation, which is performed using a dynamics-informed health indicator derived from GAR parameters and a simple linear Wiener process model. Training for the platform relies on a limited set of experimental vibration signals from the physical drivetrain, augmented with high-fidelity multibody dynamics simulations and surrogate-model realizations to ensure coverage of the full space of OCs and fault scenarios. Its performance is validated on hundreds of inspection experiments using confusion matrices, ROC curves, and metric-based plots, while the decision fusion scheme significantly strengthens diagnostic reliability across the CM stages. The results demonstrate near-perfect fault detection (99.8%), 97.8% accuracy in fault type identification, and over 96% in severity characterization. Moreover, the method yields reliable early-stage RUL estimates for the outer gear of the drivetrain, with normalized errors < 20% and consistently narrow confidence bounds, which confirms the platform’s robustness and practicality for real-world drivetrain systems monitoring. Full article