Search Results (35,088)

This paper presents a novel hybrid algorithm for determining the optimal Phasor Measurement Units (PMU) configuration in power networks to ensure full topological and numerical observability through a multi-phase process. In the first phase, a graph-theoretic Heuristic Node Selector (HNS) is developed to rapidly establish topological observability via Core-Tree construction and node dominance evaluation. Unlike most existing studies that implicitly assume topological observability implies numerical observability, the second phase applies a Genetic Algorithm to refine and extend the initial solution from HNS, ensuring complete numerical observability while minimizing number of PMUs. This hybrid method significantly reduces the search space and improves convergence. The HNS procedure is further extended in this work to explicitly handle Zero Injection Buses (ZIB) through rule-based topological modifications, enabling a modified version of the algorithm applicable to real networks with complex structures. Real-world implementation practices from European Transmission System Operators are considered through the adoption of a “one PMU per feeder” configuration. The proposed method is validated on standard IEEE test systems and Serbian transmission networks. Results demonstrate high scalability, adaptability to various network topologies (with and without ZIB nodes), and efficient PMU allocation. Notably, the method consistently achieves high values of the System Observability Redundancy Index, indicating strong robustness and redundancy in measurement placement. Full article

Deep learning-based methods for land cover semantic change detection utilizing high-resolution, multi-temporal remote sensing imagery have emerged as a research hotspot. However, traditional CNN methods often struggle to preserve long-range spatial context information and face challenges in detecting land cover types with complex semantic change patterns in natural scenes. To address these issues, this study proposes a novel network architecture that integrates a Siamese network with differential structures and a Transformer. First, we introduce residual learning modules to improve the extraction of differential features and strengthen the representation of local features. Second, we integrate the Cross-Shaped Window (CSWin) Transformer into a differential-based Siamese network to enhance global feature extraction. To promote model training and evaluation, we propose a rural land cover change detection dataset—a high-precision dataset comprising 6 main rural land cover types. Ablation and comparative experiments were conducted on the publicly available SECOND datasets and the self-built RLCD dataset. Ablation studies on the RLCD dataset demonstrate that DSTNet achieves significant improvements over the baseline, with increases of 1.77%, 1.95%, 2.57%, and 0.92% in mIoU, Sek, Fscd, and OA. Comparative experiments on the SECOND datasets reveal that the mIoU, Sek, Fsd, and OA scores of DSTNet surpassed the second-best accuracy by 1.04%, 2.15%, 2.28%, and 0.72%. Full article

Against global challenges like climate change and biodiversity loss, sustainable development is the core orientation of engineering education transformation. Cultivating talents with interdisciplinary perspectives, systemic thinking and AI literacy is crucial for implementing the UN 2030 Sustainable Development Agenda. However, AI education focuses on seniors or graduates, with freshmen’s use of AI acting as “cognitive partners” for knowledge construction and complex problem-solving understudied, constraining AI’s potential in fostering early systemic thinking. We present a novel teaching practice integrating generative AI into an “AI-Environmental System Analysis” module, with Sousa chinensis habitat conservation as the case. Using a design-based research paradigm, we evaluated 24 student groups via system analysis briefs, AI usage reflections and course assessment data. Results show that the module effectively guided students to establish preliminary system analysis frameworks, with over 70% of groups identifying complex interactions among environmental factors. Students’ AI applications ranged from information retrieval to scenario simulation, initially forming systemic thinking and responsible AI literacy for sustainable development. This study provides a replicable paradigm for integrating AI and sustainable development education, clarifies the key role of structured instructional scaffolding, and enriches sustainable development-oriented engineering education pathways. Full article

Axial postural abnormalities in Parkinson’s Disease (PD) are traditionally assessed using clinical rating scales, although picture-based assessment is considered the gold standard. This study evaluates the reliability and clinical relevance of two markerless body-tracking frameworks, the RGB-D-based Microsoft Azure Kinect (providing the reference KIN_3D model) and the RGB-only Google MediaPipe Pose (MP), using a synchronous dual-camera setup. Forty PD patients performed a 60 s static standing task. We compared KIN_3D with three MP models (at different complexity levels) across horizontal, vertical, sagittal, and 3D joint angles. Results show that lower-complexity MP models achieved high congruence with KIN_3D for trunk and shoulder alignment (ρ > 0.75), while the lateral view significantly improved tracking of sagittal angles (ρ ≥ 0.72). Conversely, the high-complexity model introduced significant skeletal distortions. Clinically, several angular parameters emerged as robust metrics for postural assessment and global motor impairments, while sagittal angles correlated with motor complications. Unexpectedly, a more upright frontal alignment was associated with greater freezing of gait severity, suggesting that static postural metrics may serve as proxies for dynamic gait performance. In addition, both RGB-only and RGB-D frameworks effectively discriminated between postural severity clusters. While the higher-complexity MP model should be avoided due to inaccurate 3D reconstructions, our findings demonstrate that low- and medium-complexity MP models represent a reliable alternative to RGB-D sensors for objective postural assessment in PD, facilitating the widespread application of objective posture measurements in clinical contexts. Full article

Landslide susceptibility mapping (LSM) is widely used for identifying potential landslide-prone areas. However, many existing approaches rely on global models that assume spatial stationarity, which limits their ability to capture spatially heterogeneous relationships in complex mountainous regions. To address this issue, this study improved landslide susceptibility evaluation by accounting for spatial heterogeneity using a Geographically Weighted Random Forest (GWRF) model. By allowing the influence of conditioning factors to vary spatially, the proposed method provides a more adaptive representation of landslide susceptibility compared to conventional global models. The GWRF-based evaluation results were compared with those obtained from Random Forest (RF) and XGBoost models to examine relative performance. The study was conducted in Yingjiang County, a landslide-prone mountainous area, using multiple landslide conditioning factors, including topographic and anthropogenic variables such as slope and distance to roads. Landslide susceptibility maps were generated, and the evaluation results were supported by InSAR-derived deformation data, field investigations, and UAV observations. The results indicate that the GWRF model achieved superior overall susceptibility evaluation performance compared to the RF and XGBoost models, with an AUC value of 0.922. Furthermore, compared to global models, the GWRF model revealed more detailed spatial patterns of landslide susceptibility, particularly in high-susceptibility zones. Areas classified as highly susceptible by the GWRF model also demonstrated greater consistency with observed deformation features. These findings highlight the importance of considering spatial heterogeneity in landslide susceptibility evaluation and demonstrate that the proposed GWRF approach is applicable for regional-scale susceptibility assessment in complex mountainous environments. Full article

The analysis of Synthetic Aperture Radar (SAR) imagery is essential to modern remote sensing, with applications in disaster management, agricultural monitoring, and military surveillance. A significant challenge is that the complex and noisy nature of SAR data severely limits the performance of traditional machine learning (TML) methods, leading to high error rates. In contrast, deep learning (DL) has recently proven highly effective at addressing these limitations. This study provides a comprehensive review of recent DL advances applied to SAR image despeckling, segmentation, classification, and detection. It evaluates widely adopted models, examines the potential of underutilized ones like GANs and GNNs, and compiles available datasets to support researchers. This review concludes by outlining key challenges and proposing future research directions to guide continued progress in SAR image analysis. Full article

This study presents a hybrid digital twin system designed for real-time indoor air quality (IAQ) monitoring and filtration optimization within a residential environment. Using a network of low-cost sensors, physics-based simulations, and machine learning models, the system dynamically replicates the indoor environment to enable continuous assessment and optimization of key pollutants, including particulate matter, volatile organic compounds, and carbon dioxide. The system architecture integrates mass balance and decay models, computational fluid dynamics simulations, regression models, and neural network algorithms, all evaluated under both filtering and non-filtering conditions. A graphical user interface allows users to interact with the system, test air purifier placements, and visualize air quality dynamics in real time. The results demonstrate that, within this system, simpler models, such as linear regression, outperform more complex architectures under data-limited conditions, achieving test-set coefficients of determination ranging from 0.97 to 0.99 across multiple IAQ parameters. At the same time, the hybrid modelling approach enhances interpretability and robustness. Overall, this digital twin system contributes to smart building management by offering a scalable, interpretable, and cost-effective solution for proactive IAQ control and personalized decision-making. Full article

In the context of increasingly complex offshore drilling operations and stricter environmental regulations, the efficient handling and volume reduction of drilling cuttings has emerged as a crucial focus in the advancement of solids control equipment. “Airflow-assisted screening” is a technique that uses directed air currents to enhance the separation of solid cuttings from drilling fluid on a shaker screen, thereby improving dewatering efficiency and reducing waste volume during drilling. This study proposes and designs novel negative-pressure suction-type cuttings reduction equipment by integrating this technology with screw conveying principles. The system features a compact, vacuum-generator-centered design that integrates suction and screening. Key components were optimized, and a monitoring scheme was implemented for real-time performance evaluation. In the mechanism analysis, the relationship between inlet pressure, geometric parameters, and suction performance was explored based on Bernoulli’s principle and Laval nozzle characteristics, and internal flow field characteristics were revealed through computational fluid dynamics (CFDs) simulations. In the experimental section, a prototype system and testing platform were constructed to evaluate the effects of inlet pressure and screen mesh configurations on suction and screening performance. The results indicate that the system achieved optimal performance at an inlet pressure of 400 kPa with a 100-mesh screen, reaching a cuttings reduction efficiency of 9.225%. This study effectively validates the theoretical and simulation findings, providing technical support for the application of this equipment in complex drilling environments and demonstrating strong potential for practical implementation. Full article

This research presents a large-scale thematic analysis of 66,002 Slovak university thesis abstracts, aimed at identifying, categorizing, and visualizing research trends across multiple academic disciplines. Using BERTopic for unsupervised topic modeling with K-Means clustering, 3000 distinct thematic clusters were extracted through rigorous coherence optimization, with each topic characterized by representative keywords derived from class-based TF-IDF weighting. Text embeddings were generated using SlovakBERT-STS, a domain-adapted Slovak BERT model fine-tuned for semantic textual similarity, producing 768-dimensional vectors that enable precise computation of cosine similarity between topics, resulting in a 3000 × 3000 topic similarity matrix. The optimal topic count was determined through systematic evaluation of K values ranging from 1000 to 10,000, with K = 3000 identified as the optimal configuration based on coherence elbow analysis, yielding a mean coherence score of 0.433. Thematic relationships were visualized through Multidimensional Scaling (MDS) projection to 3-D space, where convex hull geometries reveal semantic boundaries and topic separability. The methodology incorporates dynamic stopword filtering, Stanza-based lemmatization for Slovak morphology, and UMAP dimensionality reduction, achieving a balanced distribution of approximately 22 abstracts per topic. Results demonstrate that fine-grained topic models with 3000 clusters can extract meaningful semantic structure from multi-domain, morphologically complex Slovak academic corpora, despite inherent coherence constraints. The reproducible pipeline provides a framework for large-scale topic discovery, coherence-driven optimization, and geometric visualization of thematic relationships in academic text collections. Full article

The regulatory framework governing EU electricity markets is highly complex, fragmented across multiple normative acts and sensitive to citation accuracy and contextual completeness. While Large Language Models (LLMs) offer promising capabilities for regulatory question answering (QA), their tendency to hallucinate legal references and omit critical conditions makes them unreliable for compliance-sensitive domains. This paper presents the design of a domain-specific Retrieval-Augmented Generation (RAG) system for EU electricity market regulations, explicitly engineered to deliver source-grounded, traceable and low-hallucination answers. The answering component is based on Google’s gemini-2.5-flash model. The Open AI’s gpt-4o-mini model is responsible for both relevant document selection before building the RAG prompt and playing the judge LLM role for Retrieval Augmented Generation Assessment (RAGAS) evaluation. We build a legal corpus comprising multiple core EU regulatory acts related to REMIT and market operation and propose a regulatory QA architecture that integrates: (i) three chunking strategies (article-based, structure-aware, sliding window), (ii) two embedding models and (iii) a novel LLM-based document selection agent that restricts retrieval to the most relevant normative acts before vector search, improving contextual focus and retrieval precision. Using a fixed benchmark of regulatory questions and a reproducible evaluation protocol, we quantitatively assess system performance with RAGAS metrics and classical information-retrieval measures. While all configurations achieve strong faithfulness (up to 0.96), answer relevancy varies substantially with embedding and chunking choices. The findings confirm that retrieval engineering, particularly embedding selection, chunking strategy and pre-retrieval document filtering, has a high impact for building reliable regulatory AI systems. The sliding window strategy combined with bge-small-en-v1.5 delivered the strongest rank-sensitive retrieval performance, achieving the highest Precision@10 and NDCG@10. In contrast, article-level chunking with the same model yielded a modest improvement in Recall@10, indicating a clear trade-off between recall and precision-oriented ranking quality in legal corpora. Full article

Metaheuristic optimization algorithms are widely employed to address complex nonlinear and multimodal optimization problems due to their flexibility and strong global search capability. However, the original Connected Banking System Optimizer (CBSO) still exhibits several inherent limitations when handling high-dimensional and highly complex search spaces, including excessive dependence on single global-best guidance, rapid loss of population diversity, weak exploitation ability in later iterations, and inefficient boundary handling. These deficiencies often lead to premature convergence and unstable optimization performance. To overcome these drawbacks, this paper proposes a Multi-Strategy Enhanced Connected Banking System Optimizer (MSECBSO) by systematically enhancing the CBSO framework through multiple complementary mechanisms. First, a multi-elite cooperative guidance strategy is introduced to aggregate information from several high-quality individuals, thereby mitigating search-direction bias and improving population diversity. Second, an embedded differential evolution search strategy is incorporated to strengthen local exploitation accuracy and enhance the ability to escape from local optima. Third, a soft boundary rebound mechanism is designed to replace rigid boundary truncation, improving search stability and preventing boundary aggregation. The proposed MSECBSO is extensively evaluated on the CEC2017 and CEC2022 benchmark suites under different dimensional settings and is statistically compared with nine state-of-the-art metaheuristic algorithms. Experimental results demonstrate that MSECBSO achieves superior convergence accuracy, robustness, and stability across unimodal, multimodal, hybrid, and composition functions. In terms of computational complexity, MSECBSO retains the same order of time complexity as the original CBSO, namely O(N×D×T), while introducing only a marginal increase in constant computational overhead. The space complexity remains O(N×D), indicating good scalability for high-dimensional optimization problems. Furthermore, MSECBSO is applied to corporate bankruptcy forecasting by optimizing the hyperparameters of a K-nearest neighbors (KNN) classifier. The resulting MSECBSO-KNN model achieves higher prediction accuracy and stronger stability than competing optimization-based KNN models, confirming the effectiveness and practical applicability of the proposed algorithm in real-world classification tasks. Full article

Background and Clinical Significance: Radiocarpal fracture dislocations (RCFDs) are rare injuries of the wrist, while open RCFDs represent a small subgroup of these injuries. Limited data exists regarding the optimal method for their management. Our study’s objective is to present a rare case of an open (Gustilo–Anderson type II) dorsal radiocarpal dislocation in combination with fracture of the radial and ulnar styloid and neurologic deficits (superficial radial, median and ulnar nerve), which was treated with external fixation and Kirshner wire pinning. External fixation and Kirshner wire pinning could be a viable surgical option for complicated open RCFD. Case Presentation: Adequate reduction and ligamentotaxis using an external fixation were achieved, while the radial styloid fracture and the distal radioulnar joint (DRJ) were stabilized with Kirshner wires. Postoperative radiographs and clinical evaluation confirmed satisfactory reduction in the right wrist, without signs of intercarpal instability. Total nerve recovery was observed 6 months postoperatively and the patient was able to return to his previous occupation. At the final follow-up (4 years), the Visual Analogue Scale score was 1/10 and the Quick Dash score was 11/100 with good range of motion (flexion: 0–75°, extension: 0–70°, pronation: 0–80°, supination: 0–80°) of the affected wrist, although progressive wrist arthritis and ulnar migration was seen in the plain X-rays. Conclusions: Surgical treatment of RCFDs is required for complex or unstable fractures/dislocations to avoid possible complications, such as intercarpal instability. Full article

The advancement of Industry 5.0 hinges on the deep integration of artificial intelligence (AI) with domain expertise to foster sustainable industrial development. This study proposes a deep learning-based surrogate modeling framework that integrates reservoir production requirements with AI technologies, providing intelligent decision support for production optimization and enhanced efficiency. To evaluate AI’s effectiveness in complex industrial scenarios, we conduct an integrated analysis encompassing model construction, dynamic prediction, and production optimization using a real-world oilfield case. This oilfield features a dynamically increasing number of wells and requires dynamic adjustments to injection–production relationships. To address this challenge, we enhance the Embed-to-Control model by improving the nonlinear representation capability within its decoder structure. Subsequently, we construct a high-fidelity dataset containing 300 samples for model training and testing. The experimental results demonstrate that the proposed improved model achieves a high accuracy in predicting key state variables (pressure and saturation) and oil production. Regarding computational efficiency, a single model run requires only approximately 17.3 s, achieving an over 200× speedup relative to traditional numerical simulators. Finally, we coupled the trained surrogate model with the particle swarm optimization algorithm to optimize the injection well control strategy. The optimized scheme increases daily oil production by 13.84%, boosting economic benefits. This study demonstrates a practical technological pathway to accelerate the oil and gas industry’s transition toward Industry 5.0. Full article

Background/Objectives: Hanging is the most common method of suicide in most countries worldwide. It is characterized by high lethality, technical simplicity, and typical autopsy findings. Autopsy plays a crucial role in determining the cause and mechanism of death. While external injuries are relatively consistent, internal findings show considerable variability in the literature. The aim of this prospective study was to analyze 660 cases of suicidal hanging over a ten-year period, focusing on the occurrence of forensically relevant internal autopsy findings. Methods: The study was conducted at the Department of Forensic Medicine in Bratislava between 2015 and 2024. All cases underwent standardized complete autopsy, including histology, toxicology, and analysis of death circumstances. Recently reported thoracic aortic adventitial hemorrhages described in 2024 were evaluated only in a targeted subset of cases examined between July and December 2024. Statistical evaluation was performed using the chi-square test to identify associations between internal findings and suspension type, knot location, age, sex, and body weight. Results: The argent line was present in 61.1% of cases, most frequently with posterior knot placement and complete suspension. Neck muscle hemorrhages occurred in 53.8%, predominantly at the periosteal-clavicular attachment of the sternocleidomastoid muscle, with higher incidence in complete and anterior suspension. Amussat’s sign was observed in 10.2% of cases, and Etienne-Martin’s sign in 1.1%. Fractures of the laryngo-hyoid complex were present in 49.7%, mainly in cases with complete suspension and posterior knot location. Cervical spine injuries were detected in 2.6%, predominantly in older males and with anterior knot placement. Simon’s hemorrhages occurred in 35.2%, mainly in younger individuals and complete suspension. Hemorrhages in the intestinal wall were detected in 7.4%, and rectal hemorrhages in 1.1% of cases. In the targeted 2024 subset, no thoracic aortic adventitial hemorrhages were identified. Conclusions: The findings suggest the forensic relevance of several internal findings associated with hanging, while emphasizing that the results were obtained using a uniform and consistently applied autopsy protocol. They also indicate the need for further research, particularly regarding recently reported adventitial hemorrhages of the thoracic aorta, which were assessed only in a limited subset of cases during the final months of the study and were not identified in our material. Full article

Financial document understanding remains a critical challenge for Large Language Models, primarily due to the complex interplay between narrative text and structured numerical tables. Existing Retrieval-Augmented Generation (RAG) systems often treat these modalities in isolation, leading to significant failures in tasks requiring joint reasoning. This study introduces HierFinRAG, a novel hierarchical multimodal framework designed to unify tabular and textual data processing. Our approach employs a Table-Text Graph Neural Network (TTGNN) to explicitly model semantic and structural dependencies between table cells and corresponding text, coupled with a Symbolic–Neural Fusion module that routes queries between a neural generator and a symbolic calculator for precise arithmetic operations. We evaluate the system on the FinQA and FinanceBench datasets, comparing performance against strong baselines including Vanilla RAG and GPT-4o with Code Interpreter. Results demonstrate that HierFinRAG achieves an Exact Match score of 82.5% on FinQA, surpassing the best baseline by 6.5 percentage points, while maintaining a 3.5× faster inference latency than agentic approaches. These findings indicate that integrating hierarchical structural awareness with hybrid reasoning significantly enhances the accuracy and interpretability of financial artificial intelligence systems. Full article