This paper presents an new approach for detecting in the electrical power system of satellites operating in Low Earth Orbit (LEO) without an Attitude Determination and Control Subsystem (ADCS). Components of these systems are prone to faults, such as line-to-line faults in the photovoltaic subsystem, open circuits, and short circuits in the DC-to-DC converter, as well as ground faults in batteries. In the previous research has largely focused on detecting faults in each components, such as photovoltaic arrays or converter systems, therefore, has been limited attention given to whole electrical power system of satellite as a whole system. Our approach addresses this gap by utilizing a Multi-Layer Perceptron (MLP) neural network model, which leverages input data such as solar radiation and surface temperature to predict current and load outputs. These machine learning techniques that classifiy use different approaches like Principal Component Analysis (PCA) and K-Nearest Neighbors (KNN), to classify faults effectively. The model presented achieves over 99% accuracy in identifying faults across multiple subsystems, marking a notable advancement from previous approaches by offering a complete diagnostic solution for the entire satellite power system. This thorough method boosts system reliability and helps lower the chances of mission failure
The increasing proliferation of vending machines in public and commercial environments has placed a growing emphasis on operational efficiency and customer satisfaction. Traditional maintenance approaches either reactive or time-based preventive are limited in their ability to preempt machine failures, leading to unplanned downtimes and elevated service costs. This research presents a novel predictive maintenance framework tailored for vending machines by leveraging Internet of Things (IoT) sensors and machine learning (ML) algorithms. The proposed system continuously monitors machine components and operating conditions in real time and applies predictive models to forecast failures before they occur. This enables timely maintenance scheduling, minimizing downtime and extending machine lifespan. The framework was validated through simulated fault data and performance evaluation using classification algorithms. Results show a significant improvement in early fault detection and a reduction in redundant service interventions. The findings indicate that predictive maintenance systems, when integrated into vending infrastructure, can transform operational efficiency and service reliability.
Detecting anomalies in datasets is a longstanding problem in machine learning. In this context, anomalies are defined as a sample that significantly deviates from the remaining data. Meanwhile, optimal transport (OT) is a field of mathematics concerned with the transportation, between two probability measures, at least effort. In classical OT, the optimal transportation strategy of a measure to itself is the identity. In this paper, we tackle anomaly detection by forcing samples to displace its mass, while keeping the least effort objective. We call this new transportation problem Mass Repulsing Optimal Transport (MROT). Naturally, samples lying in low density regions of space will be forced to displace mass very far, incurring a higher transportation cost. We use these concepts to design a new anomaly score. Through a series of experiments in existing benchmarks, and fault detection problems, we show that our algorithm improves over existing methods.
Evaluating anomaly detection in multivariate time series (MTS) requires careful consideration of temporal dependencies, particularly when detecting subsequence anomalies common in fault detection scenarios. While time series cross-validation (TSCV) techniques aim to preserve temporal ordering during model evaluation, their impact on classifier performance remains underexplored. This study systematically investigates the effect of TSCV strategy on the precision-recall characteristics of classifiers trained to detect fault-like anomalies in MTS datasets. We compare walk-forward (WF) and sliding window (SW) methods across a range of validation partition configurations and classifier types, including shallow learners and deep learning (DL) classifiers. Results show that SW consistently yields higher median AUC-PR scores and reduced fold-to-fold performance variance, particularly for deep architectures sensitive to localized temporal continuity. Furthermore, we find that classifier generalization is sensitive to the number and structure of temporal partitions, with overlapping windows preserving fault signatures more effectively at lower fold counts. A classifier-level stratified analysis reveals that certain algorithms, such as random forests (RF), maintain stable performance across validation schemes, whereas others exhibit marked sensitivity. This study demonstrates that TSCV design in benchmarking anomaly detection models on streaming time series and provide guidance for selecting evaluation strategies in temporally structured learning environments.
Accurate state estimation is essential for monitoring and controlling nonlinear chemical reactors, such as continuous stirred-tank reactors (CSTRs), where limited sensor coverage and process uncertainties hinder real-time observability. This paper introduces a novel multi-observer switching framework that combines several advanced estimators: Extended Luenberger Observer (ELO), Extended Kalman Filter (EKF), Unscented Kalman Filter (UKF), Quadrature Kalman Filter (QKF), and Particle Filter (PF), operating in parallel. At each sampling instant, a cost function based on the $L_1$ norm and Kullback-Leibler divergence selects the observer, yielding the best agreement with available measurements. The proposed architecture is validated through simulations of both linearized and fully nonlinear CSTR models with up to three reactors in series. Results show that the switching strategy significantly reduces estimation errors compared to single-observer approaches, especially under partial observability and parametric uncertainty. Despite its modular design, the framework remains computationally tractable, making it suitable for real-time industrial applications such as fault detection and model-based predictive control.
Lithium-ion batteries are the enabling power source for transportation electrification. However, in real-world applications, they remain vulnerable to internal short circuits (ISCs) and the consequential risk of thermal runaway (TR). Toward addressing the challenge of ISCs and TR, we undertake a systematic study that extends from dynamic modeling to fault detection in this paper. First, we develop {\em BattBee}, the first equivalent circuit model to specifically describe the onset of ISCs and the evolution of subsequently induced TR. Drawing upon electrochemical modeling, the model can simulate ISCs at different severity levels and predict their impact on the initiation and progression of TR events. With the physics-inspired design, this model offers strong physical interpretability and predictive accuracy, while maintaining structural simplicity to allow fast computation. Then, building upon the BattBee model, we develop fault detection observers and derive detection criteria together with decision-making logics to identify the occurrence and emergence of ISC and TR events. This detection approach is principled in design and fast in computation, lending itself to practical applications. Validation based on simulations and experimental data demonstrates the effectiveness of both the BattBee model and the ISC/TR detection approach. The research outcomes underscore this study's potential for real-world battery safety risk management.
Future electrical grids will require new ways to identify faults as inverters are not capable of supplying large fault currents to support existing fault detection methods and because distributed resources may feed faults from the edge of the grid. This paper proposes the use of real-time system identification for online power-system fault detection. Specifically, we implement Recursive ARX (rARX) system identification on a grid-connected inverter. Experiments demonstrate that the proposed rARX method is able to both detect large faults quickly, and distinguish between high-impedance faults and large load increases. These results indicate that rARX grid-edge fault detection is a promising research direction for improving the reliability and safety of modern electric grids.
This paper presents a joint system modeling approach for fault simulation of all-electric auxiliary power unit (APU), integrating starter/generator turn-to-turn short circuit (TTSC) faults with gas generator gas-path faults.To address challenges in electromechanical coupling, simulation precision and computational efficiency balance, we propose a multi-rate continuous-discrete hybrid simulation architecture. This architecture treats the starter/generator as a continuous system with variable step size in Simulink, while modeling the gas generator as a discrete system with fixed step size in a dynamic-link library (DLL) environment. For the starter/generator fault modeling, a multi-loop approach is deployed to accurately simulate TTSC faults. For the gas generator, we develop an improved GasTurb-DLL modeling method (IGDM) that enhances uncertainty modeling, state-space representation, and tool chain compatibility. Finally, the proposed methodology above was implemented in a case study based on the APS5000 all-electric APU structure and parameters. Model validation was conducted by comparing simulation results--covering steady-state, transients, healthy, and fault conditions--with reference data from third-party software and literature. The close agreement confirms both the model's accuracy and the effectiveness of our modeling methodology. This work establishes a modeling foundation for investigating the opportunities and challenges in fault detection and isolation (FDI) brought by the all electrification of the APU, including joint fault estimation and diagnosis, coupled electromechanical fault characteristics.
In this paper we address the problem of online detection of inter-turn short-circuit faults (ITSCFs) that occur in permanent magnet synchronous motors (PMSMs). We propose two solutions to this problem: (i) a very simple linear observer and (ii) a generalized parameter estimation based observer, that incorporates a high performance estimator -- with both observers detecting the short-circuit current and the fault intensity. Although the first solution guarantees the detection of the fault exponentially fast, the rate of convergence is fully determined by the motor parameters that, in some cases, may be too slow. The second observer, on the other hand, ensures finite convergence time under the weakest assumption of interval excitation. To make the observers adaptive, we develop a parameter estimator that, in the case of isotropic PMSMs, estimates on-line (exponentially fast) the resistance and inductance of the motor. It should be underscored that, in contrast with existing observers (including the widely popular Kalman filter) that provide indirect information of the fault current, our observers provide explicit one -- namely the amplitude of the fault current. The performance of both observers, in their linear and generalized parameter estimation-based versions, is illustrated with realistic simulation studies.
This paper addresses the challenges of fault prediction and delayed response in distributed systems by proposing an intelligent prediction method based on temporal feature learning. The method takes multi-dimensional performance metric sequences as input. We use a Gated Recurrent Unit (GRU) to model the evolution of system states over time. An attention mechanism is then applied to enhance key temporal segments, improving the model's ability to identify potential faults. On this basis, a feedforward neural network is designed to perform the final classification, enabling early warning of system failures. To validate the effectiveness of the proposed approach, comparative experiments and ablation analyses were conducted using data from a large-scale real-world cloud system. The experimental results show that the model outperforms various mainstream time-series models in terms of Accuracy, F1-Score, and AUC. This demonstrates strong prediction capability and stability. Furthermore, the loss function curve confirms the convergence and reliability of the training process. It indicates that the proposed method effectively learns system behavior patterns and achieves efficient fault detection.
Recent advancements in language modeling tasks have been driven by architectures such as Transformers and, more recently, by Selective State Space Models (SSMs). In this paper, we introduce an alternative selection mechanism inspired by control theory methodologies. Specifically, we propose a novel residual generator for selection, drawing an analogy to fault detection strategies in Linear Time-Invariant (LTI) systems. Unlike Mamba, which utilizes Linear Time-Varying (LTV) systems, our approach combines multiple LTI systems, preserving their beneficial properties during training while achieving comparable selectivity. To evaluate the effectiveness of the proposed architecture, we test its performance on synthetic tasks. While these tasks are not inherently critical, they serve as benchmarks to test the selectivity properties of different cores architecture. This work highlights the potential of integrating theoretical insights with experimental advancements, offering a complementary perspective to deep learning innovations at the intersection of control theory and machine learning.
Germany's transition to a renewable energy-based power system is reshaping grid operations, requiring advanced monitoring and control to manage decentralized generation. Machine learning (ML) has emerged as a powerful tool for power system protection, particularly for fault detection (FD) and fault line identification (FLI) in transmission grids. However, ML model reliability depends on data quality and availability. Data sparsity resulting from sensor failures, communication disruptions, or reduced sampling rates poses a challenge to ML-based FD and FLI. Yet, its impact has not been systematically validated prior to this work. In response, we propose a framework to assess the impact of data sparsity on ML-based FD and FLI performance. We simulate realistic data sparsity scenarios, evaluate their impact, derive quantitative insights, and demonstrate the effectiveness of this evaluation strategy by applying it to an existing ML-based framework. Results show the ML model remains robust for FD, maintaining an F1-score of 0.999 $\pm$ 0.000 even after a 50x data reduction. In contrast, FLI is more sensitive, with performance decreasing by 55.61% for missing voltage measurements and 9.73% due to communication failures at critical network points. These findings offer actionable insights for optimizing ML models for real-world grid protection. This enables more efficient FD and supports targeted improvements in FLI.
Large Language Models (LLMs) have recently emerged as promising tools for automated unit test generation. We introduce a hybrid framework called EvoGPT that integrates LLM-based test generation with evolutionary search techniques to create diverse, fault-revealing unit tests. Unit tests are initially generated with diverse temperature sampling to maximize behavioral and test suite diversity, followed by a generation-repair loop and coverage-guided assertion enhancement. The resulting test suites are evolved using genetic algorithms, guided by a fitness function prioritizing mutation score over traditional coverage metrics. This design emphasizes the primary objective of unit testing-fault detection. Evaluated on multiple open-source Java projects, EvoGPT achieves an average improvement of 10% in both code coverage and mutation score compared to LLMs and traditional search-based software testing baselines. These results demonstrate that combining LLM-driven diversity, targeted repair, and evolutionary optimization produces more effective and resilient test suites.
With the rapid development of cloud computing systems and the increasing complexity of their infrastructure, intelligent mechanisms to detect and mitigate failures in real time are becoming increasingly important. Traditional methods of failure detection are often difficult to cope with the scale and dynamics of modern cloud environments. In this study, we propose a novel AI framework based on Massive Language Model (LLM) for intelligent fault detection and self-healing mechanisms in cloud systems. The model combines existing machine learning fault detection algorithms with LLM's natural language understanding capabilities to process and parse system logs, error reports, and real-time data streams through semantic context. The method adopts a multi-level architecture, combined with supervised learning for fault classification and unsupervised learning for anomaly detection, so that the system can predict potential failures before they occur and automatically trigger the self-healing mechanism. Experimental results show that the proposed model is significantly better than the traditional fault detection system in terms of fault detection accuracy, system downtime reduction and recovery speed.
There is currently a paradigm shift in several power system monitoring applications, such as incipient fault detection and monitoring inverter-based resources, to transition from traditional phasor analytics to more informative waveform analytics. This paper contributes to this transition by developing a novel approach to modeling voltage and current waveform measurements using implicit neural representations (INRs). INRs are continuous function approximators that are recently used in vision and signal processing. The proposed INR models are specifically designed to meet the requirements of waveform analytics in power systems, such as by using sinusoidal activation functions that capture the periodic nature of voltage and current waveforms. We also propose extended models that can efficiently represent correlated waveforms, such as three-phase waveforms and synchro-waveforms. Real-world case studies demonstrate the effectiveness of the proposed INR models in terms of accuracy (<1-2% MSE) and model size (4-6x compression). We also investigate the application of INR models in oscillation monitoring, for single mode oscillations and dual mode modulated oscillations.
Modern robots face challenges shared by humans, where machines must learn multiple sensorimotor skills and express them adaptively. Equipping robots with a human-like memory of how it feels to do multiple stereotypical movements can make robots more aware of normal operational states and help develop self-preserving safer robots. Associative Skill Memories (ASMs) aim to address this by linking movement primitives to sensory feedback, but existing implementations rely on hard-coded libraries of individual skills. A key unresolved problem is how a single neural network can learn a repertoire of skills while enabling fault detection and context-aware execution. Here we introduce Neural Associative Skill Memories (ASMs), a framework that utilises self-supervised predictive coding for temporal prediction to unify skill learning and expression, using biologically plausible learning rules. Unlike traditional ASMs which require explicit skill selection, Neural ASMs implicitly recognize and express skills through contextual inference, enabling fault detection across learned behaviours without an explicit skill selection mechanism. Compared to recurrent neural networks trained via backpropagation through time, our model achieves comparable qualitative performance in skill memory expression while using local learning rules and predicts a biologically relevant speed-accuracy trade-off during skill memory expression. This work advances the field of neurorobotics by demonstrating how predictive coding principles can model adaptive robot control and human motor preparation. By unifying fault detection, reactive control, skill memorisation and expression into a single energy-based architecture, Neural ASMs contribute to safer robotics and provide a computational lens to study biological sensorimotor learning.
The anomaly detection literature is abundant with offline methods, which require repeated access to data in memory, and impose impractical assumptions when applied to a streaming context. Existing online anomaly detection methods also generally fail to address these constraints, resorting to periodic retraining to adapt to the online context. We propose Online-iForest, a novel method explicitly designed for streaming conditions that seamlessly tracks the data generating process as it evolves over time. Experimental validation on real-world datasets demonstrated that Online-iForest is on par with online alternatives and closely rivals state-of-the-art offline anomaly detection techniques that undergo periodic retraining. Notably, Online-iForest consistently outperforms all competitors in terms of efficiency, making it a promising solution in applications where fast identification of anomalies is of primary importance such as cybersecurity, fraud and fault detection.
A fault detection method for power conversion circuits using thermal images and a convolutional autoencoder is presented. The autoencoder is trained on thermal images captured from a commercial power module at randomly varied load currents and augmented image2 generated through image processing techniques such as resizing, rotation, perspective transformation, and bright and contrast adjustment. Since the autoencoder is trained to output images identical to input only for normal samples, it reconstructs images similar to normal ones even when the input images containing faults. A small heater is attached to the circuit board to simulate a fault on a power module, and then thermal images were captured from different angles and positions, as well as various load currents to test the trained autoencoder model. The areas under the curve (AUC) were obtained to evaluate the proposed method. The results show the autoencoder model can detect anomalies with 100% accuracy under given conditions. The influence of hyperparameters such as the number of convolutional layers and image augmentation conditions on anomaly detection accuracy was also investigated.
Large Language Models (LLMs) are increasingly deployed in various applications, raising critical concerns about fairness and potential biases in their outputs. This paper explores the prioritization of metamorphic relations (MRs) in metamorphic testing as a strategy to efficiently detect fairness issues within LLMs. Given the exponential growth of possible test cases, exhaustive testing is impractical; therefore, prioritizing MRs based on their effectiveness in detecting fairness violations is crucial. We apply a sentence diversity-based approach to compute and rank MRs to optimize fault detection. Experimental results demonstrate that our proposed prioritization approach improves fault detection rates by 22% compared to random prioritization and 12% compared to distance-based prioritization, while reducing the time to the first failure by 15% and 8%, respectively. Furthermore, our approach performs within 5% of fault-based prioritization in effectiveness, while significantly reducing the computational cost associated with fault labeling. These results validate the effectiveness of diversity-based MR prioritization in enhancing fairness testing for LLMs.
Accurate diagnosis of power transformer faults is essential for ensuring the stability and safety of electrical power systems. This study presents a comparative analysis of conventional machine learning (ML) algorithms and deep learning (DL) algorithms for fault classification of power transformers. Using a condition-monitored dataset spanning 10 months, various gas concentration features were normalized and used to train five ML classifiers: Support Vector Machine (SVM), k-Nearest Neighbors (KNN), Random Forest (RF), XGBoost, and Artificial Neural Network (ANN). In addition, four DL models were evaluated: Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), One-Dimensional Convolutional Neural Network (1D-CNN), and TabNet. Experimental results show that both ML and DL approaches performed comparably. The RF model achieved the highest ML accuracy at 86.82%, while the 1D-CNN model attained a close 86.30%.