Search Results (4,215)

Electric vehicle (EV) charging has become a key factor in grid integration, impact analysis, and the development of intelligent charging strategies. However, the rapid rise in EV adoption poses challenges for charging infrastructure and grid stability due to the inherently variable and uncertain charging behavior. Limited access to high-resolution, location-specific data further hinders accurate modeling, emphasizing the need for reliable, privacy-preserving tools to forecast EV-related grid impacts. This study introduces a comprehensive methodology to emulate real-world EV charging behavior using a real-time simulation environment. A physics-based EV charger model was developed on the Typhoon HIL platform, incorporating detailed electrical dynamics and control logic representative of commercial chargers. Simulation outputs, including active power consumption and state-of-charge evolution, were validated against field data captured via phasor measurement units, showing strong alignment across all charging phases, including SOC-dependent current transitions. Quantitative validation yielded an MAE of 0.14 and an RMSE of 0.36, confirming the model’s high accuracy. The study also reflects practical BMS strategies, such as early charging termination near 97% SOC to preserve battery health. Overall, the proposed real-time framework provides a high-fidelity platform for analyzing grid-integrated EV behavior, testing smart charging controls, and enabling digital twin development for next-generation electric mobility. Full article

As the safety requirements of elevator systems continue to rise, the detection of loose bolts and the high-precision segmentation of anti-loosening lines have become critical challenges in elevator landing door inspection. Traditional manual inspection and conventional visual detection often fail to meet the requirements of high precision and robustness under real-world conditions such as oil contamination and low illumination. This paper proposes two improved algorithms for detecting loose bolts and segmenting anti-loosening lines in elevator landing doors. For small-bolt detection, we introduce the DS-EMA model, an enhanced YOLOv8 variant that integrates depthwise-separable convolutions and an Efficient Multi-scale Attention (EMA) module. The DS-EMA model achieves a 2.8 percentage point improvement in mAP over the YOLOv8n baseline on our self-collected dataset, while reducing parameters from 3.0 M to 2.8 M and maintaining real-time throughput at 126 FPS. For anti-loosening-line segmentation, we develop an improved DeepLabv3+ by adopting a MobileViT backbone, incorporating a Global Attention Mechanism (GAM) and optimizing the ASPP dilation rate. The revised model increases the mean IoU to 85.8% (a gain of 5.4 percentage points) while reducing parameters from 57.6 M to 38.5 M. Comparative experiments against mainstream lightweight models, including YOLOv5n, YOLOv6n, YOLOv7-tiny, and DeepLabv3, demonstrate that the proposed methods achieve superior accuracy while balancing efficiency and model complexity. Moreover, compared with recent lightweight variants such as YOLOv9-tiny and YOLOv11n, DS-EMA achieves comparable mAP while delivering notably higher recall, which is crucial for safety inspection. Overall, the enhanced YOLOv8 and DeepLabv3+ provide robust and efficient solutions for elevator landing door safety inspection, delivering clear practical application value. Full article

The study investigates a hybrid energy system integrating photovoltaic (PV) panels, micro-CHP units, battery storage, and thermal storage to meet the winter energy demands of a residential building in Bacău, Romania. Using real-world experimental data from amorphous, polycrystalline, and monocrystalline PV panels, C++ Model 1 simulates building energy needs and PV system performance under varying irradiance levels. The results show that PV systems alone cannot meet the total winter demand, with polycrystalline slightly outperforming monocrystalline, yet still falling short. A second computational model (C++ Model 2) simulates hybrid energy flow, demonstrating how the CHP unit and storage systems can ensure off-grid autonomy. The model dynamically manages energy between components based on daily irradiance scenarios. The findings reveal critical thresholds for PV surplus, optimal CHP sizing, and realistic battery and thermal storage needs. This paper provides a practical framework for designing efficient, data-driven hybrid solar–CHP systems for cold climates. The novelty lies in the integration of real-world PV efficiency data with a dynamic irradiance-driven simulation framework, enabling precise hybrid system sizing for winter-dominant regions. Full article

With the widespread application of artificial intelligence in real estate price prediction, model explainability has become a critical factor influencing its acceptability and trustworthiness. The Shapley value, as a classic cooperative game theory method, quantifies the average marginal contribution of each feature, ensuring global fairness in the explanation allocation. However, its focus on average fairness lacks robustness under data perturbations, model changes, and adversarial attacks. To address this limitation, this paper proposes a hybrid explainability framework that integrates the Shapley value and Least Core attribution. The framework leverages the Least Core theory by formulating a linear programming problem to minimize the maximum dissatisfaction of feature subsets, providing bottom-line fairness. Furthermore, the attributions from the Shapley value and Least Core are combined through a weighted fusion approach, where the weight acts as a tunable hyperparameter to balance the global fairness and worst-case robustness. The proposed framework is seamlessly integrated into mainstream machine learning models such as XGBoost. Empirical evaluations on real-world real estate rental data demonstrate that this hybrid attribution method not only preserves the global fairness of the Shapley value but also significantly enhances the explanation consistency and trustworthiness under various data perturbations. This study provides a new perspective for robust explainable AI in high-risk decision-making scenarios and holds promising potential for practical applications. Full article

This study explores the application of federated learning (FL) in security camera surveillance systems to overcome the structural limitations inherent in traditional centralized artificial intelligence (AI) training approaches, while simultaneously enhancing operational efficiency and data security. Conventional centralized AI models require the transmission of raw surveillance data from individual security camera units to a central server for model training, which poses significant challenges, including network congestion, a heightened risk of personal data leakage, and inadequate adaptation to localized environmental characteristics. These limitations are particularly critical in high-security environments such as military bases and government facilities, where reliability and real-time processing are paramount. In contrast, FL enables decentralized training by retaining data on local devices and sharing only model parameters with a central aggregator, thereby improving privacy preservation, reducing communication overhead, and facilitating adaptive, context-aware learning. This paper does not present a new federated learning algorithm or original experiment. Instead, it synthesizes existing research findings and applies the Analytic Hierarchy Process (AHP) to evaluate and prioritize critical factors for deploying FL in surveillance systems. By combining literature-based evidence with structured expert judgment, this study provides practical guidelines for real-world application. This paper identifies four key performance metrics—detection accuracy, false alarm rate, response time, and network load—and conducts a comparative analysis of FL and centralized AI-based approaches in the recent literature. In addition, the AHP is employed to evaluate expert survey data, quantitatively prioritizing eight critical factors for effective FL implementation. The results highlight detection accuracy and data security as the most significant concerns, indicating that FL presents a promising solution for future smart surveillance infrastructures. This research contributes to the advancement of AI-powered surveillance systems that are both high-performing and resilient under stringent privacy and operational constraints. Full article

The highly heterogeneous and irregular morphology of liver tumors presents considerable challenges for automated segmentation. To better capture complex tumor structures, this study proposes a liver tumor segmentation framework based on multi-scale deformable feature fusion and global context modeling. The method incorporates three key innovations: (1) a Deformable Large Kernel Attention (D-LKA) mechanism in the encoder to enhance adaptability to irregular tumor features, combining a large receptive field with deformable sensitivity to precisely extract tumor boundaries; (2) a Context Extraction (CE) module in the bottleneck layer to strengthen global semantic modeling and compensate for limited capacity in capturing contextual dependencies; and (3) a Dual Cross Attention (DCA) mechanism to replace traditional skip connections, enabling deep cross-scale and cross-semantic feature fusion, thereby improving feature consistency and expressiveness during decoding. The proposed framework was trained and validated on a combined LiTS and MSD Task08 dataset and further evaluated on the independent 3D-IRCADb01 dataset. Experimental results show that it surpasses several state-of-the-art segmentation models in Intersection over Union (IoU) and other metrics, achieving superior segmentation accuracy and generalization performance. Feature visualizations at both encoding and decoding stages provide intuitive insights into the model’s internal processing of tumor recognition and boundary delineation, enhancing interpretability and clinical reliability. Overall, this approach presents a novel and practical solution for robust liver tumor segmentation, demonstrating strong potential for clinical application and real-world deployment. Full article

Background: The objective of this study was to assess the efficacy and safety of faricimab in diabetic macular edema (DME) in patients who were treatment-naïve or previously treated in a real-world setting. Methods: This was a retrospective, observational, single-center study that included 105 eyes from 79 patients diagnosed with DME and treated with intravitreal faricimab between January 2024 and January 2025. Patients were categorized into two groups according to their treatment history, namely treatment-naïve eyes and eyes previously treated (switch group). Functional (best-corrected visual acuity, BCVA) and anatomical (central foveal thickness, CFT; macular volume, MV) outcomes were assessed. The safety of faricimab was evaluated from changes in intraocular pressure and the occurrence of adverse events. Results: BCVA improved significantly in both groups, with a mean gain of +0.16 in treatment-naïve eyes and +0.10 in switch eyes. The mean reduction in CFT was −53.7 µm in the naïve group and −37.8 µm in the switch group. MV decreased by −0.4 mm³ overall, with significant reductions in both groups. No adverse events were reported, confirming the safety of faricimab in routine clinical practice. Conclusions: Faricimab showed significant improvements in functional and anatomical outcomes in patients with DME, regardless of the use of previous anti-VEGF therapies. These findings support the effectiveness and safety of faricimab in a real-world clinical setting and reinforce its potential as a valuable treatment option for DME. Full article

In today’s competitive digital landscape, application usability plays a critical role in user satisfaction and retention. Negative user reviews offer valuable insights into real-world usability issues, yet traditional analysis methods often fall short in scalability and contextual understanding. This paper proposes an intelligent framework that utilizes large language models (LLMs), including GPT-4, Gemini, and BLOOM, to automate the extraction of actionable usability recommendations from negative app reviews. By applying prompting and fine-tuning techniques, the framework transforms unstructured feedback into meaningful suggestions aligned with three core usability dimensions: correctness, completeness, and satisfaction. A manually annotated dataset of Instagram negative reviews was used to evaluate model performance. Results show that GPT-4 consistently outperformed other models, achieving BLEU scores up to 0.64, ROUGE scores up to 0.80, and METEOR scores up to 0.90—demonstrating high semantic accuracy and contextual relevance in generated recommendations. Gemini and BLOOM, while improved through fine-tuning, showed significantly lower performance. This study also introduces a practical, web-based tool that enables real-time review analysis and recommendation generation, supporting data-driven, user-centered software development. These findings illustrate the potential of LLM-based frameworks to enhance software usability analysis and accelerate feedback-driven design processes. Full article

Covering problems constitute a central theme in facility location research. This study extends the classical Maximal Covering Location Problem (MCLP) and Partial Set Covering Location Problem (PSCLP) to their generalized variants, in which each demand point must be simultaneously served by multiple facilities. This generalization captures reliability requirements inherent in applications such as emergency response and robust communication networks. We first present integer programming formulations for both generalized problems, followed by equivalent reformulations that facilitate algorithmic development. Building on these, we design exact Benders decomposition algorithms that exploit structural properties of the problems to achieve enhanced scalability and computational efficiency. Computational experiments on large-scale synthetic instances with up to 200,000 demand points demonstrate that our method attains more than a threefold speedup over CPLEX. We further validate the effectiveness of the proposed approach through experiments on a real-world dataset. In addition, we compare our method with a tabu search heuristic, and the numerical results show that within a fixed time limit, our method is generally able to identify higher-quality feasible solutions. These results collectively demonstrate both the effectiveness and the practical applicability of our approach for large-scale generalized covering problems. Full article

Reliable localization is critical for robot navigation in complex indoor environments. In this paper, we propose an uncertainty-aware localization method that enhances the reliability of localization outputs without modifying the prediction model itself. This study introduces a percentile-based rejection strategy that filters out unreliable 3-degree-of-freedom pose predictions based on aleatoric and epistemic uncertainties the network estimates. We apply this approach to a multi-modal end-to-end localization that fuses RGB images and 2D LiDAR data, and we evaluate it across three real-world datasets collected using a commercialized serving robot. Experimental results show that applying stricter uncertainty thresholds consistently improves pose accuracy. Specifically, the mean position error, calculated as the average Euclidean distance between the predicted and ground-truth (x, y) coordinates, is reduced by 41.0%, 56.7%, and 69.4%, and the mean orientation error, representing the average angular deviation between the predicted and ground-truth yaw angles, is reduced by 55.6%, 65.7%, and 73.3%, when percentile thresholds of 90%, 80%, and 70% are applied, respectively. Furthermore, the rejection strategy effectively removes extreme outliers, resulting in better alignment with ground truth trajectories. To the best of our knowledge, this is the first study to quantitatively demonstrate the benefits of percentile-based uncertainty rejection in multi-modal and end-to-end localization tasks. Our approach provides a practical means to enhance the reliability and accuracy of localization systems in real-world deployments. Full article

Background: Hematopoietic cell transplantation (HCT) and chimeric antigen receptor T-cell (CAR-T) therapy have markedly improved survival in pediatric patients with hematological malignancies. However, these treatments cause profound immunosuppression, leading to significant susceptibility to vaccine-preventable diseases (VPDs), including invasive pneumococcal disease and measles. Timely and tailored immunization strategies are crucial to mitigate infectious risks in this vulnerable population. Methods: We conducted a narrative review of the English-language literature from 2000 to 2024, including clinical guidelines, surveys, and original studies, to evaluate immune reconstitution and vaccination practices in pediatric patients undergoing HCT and CAR-T therapy. Literature searches in PubMed, Scopus, and Web of Science used disease-specific, therapy-specific, and pathogen-specific terms. Data synthesis focused on vaccine schedules, immune recovery markers, and adherence challenges. Results: Profound immune deficits post-HCT and CAR-T therapy compromise both innate and adaptive immunity, often necessitating revaccination. Key factors influencing vaccine responses include time since therapy, graft source, immunosuppressive treatments, and chronic graft-versus-host disease. Although inactivated vaccines are generally safe from three to six months post-HCT, live vaccines remain contraindicated until documented immune recovery. CAR-T therapy introduces unique challenges due to prolonged B-cell aplasia and hypogammaglobulinemia, leading to delayed or reduced vaccine responses. Despite established guidelines, real-world adherence to vaccination schedules remains suboptimal, driven by institutional, logistic, and patient-related barriers. Conclusions: Effective vaccination strategies are essential for reducing infectious morbidity in pediatric HCT and CAR-T recipients. Personalized vaccine schedules, immune monitoring, and multidisciplinary coordination are critical to bridging gaps between guidelines and practice, ultimately improving long-term outcomes for immunocompromised children. Full article

The efficient purification of Pb(II)-containing wastewater is essential for safeguarding public health and maintaining the aquatic environment. In this study, novel hydrous ferric oxide (HFO) nanoparticle-embedded poly(vinylidene fluoride) (PVDF) composite adsorption membranes were developed through a simple blending method for efficient Pb(II) removal. This composite membrane (denoted as HFO-PVDF) combines the excellent selectivity of HFO nanoparticles for Pb(II) with the membrane’s advantage of easy scalability. The optimized HFO-PVDF_(1.5) membrane achieved adsorption equilibrium within 20 h and exhibited excellent adsorption capacity. Moreover, adsorption capacity markedly enhanced with increasing temperature, confirming the endothermic nature of the process. The developed HFO-PVDF membranes demonstrate significant potential for real-world wastewater treatment applications, exhibiting exceptional selectivity for Pb(II) in complex ionic matrices and could be effectively regenerated via a relatively straightforward process. Furthermore, filtration and dynamic regeneration tests demonstrated that at an initial Pb(II) concentration of 5 mg/L, the membrane operated continuously for 10–13 h before regeneration, treating up to 200 L/m² of wastewater before breakthrough, highlighting potential for cost-effective industrial wastewater treatment. This study not only demonstrates the high efficiency of the HFO-PVDF membrane for heavy metal ion removal but also provides a theoretical foundation and technical support for its practical application in water treatment. Full article

Levoglucosan and its isomers, mannosan and galactosan, are widely used atmospheric tracers of biomass combustion, and levoglucosan has been previously proposed as a potential biomarker of wood smoke exposure. This study evaluated their applicability under real-world conditions. During 14-day monitoring campaigns in both heating and non-heating seasons, daily PM_2.5 and paired urine samples were collected from adults and children in two Hungarian settlements with different heating practices. Monosaccharide anhydrides in PM_2.5 and urine were quantified by gas chromatography–mass spectrometry, while demographic, dietary, and lifestyle data were obtained via questionnaires. Ambient concentrations were substantially higher during the heating season and at the rural site, confirming the significant contribution of residential wood burning to air pollution. While urinary levoglucosan was quantifiable in >90% of samples, its isomers were often below the limit of quantification. Urinary levoglucosan concentrations did not exhibit consistent seasonal or spatial patterns and were not associated with ambient levels. Instead, an unexplained background more likely influenced by certain demographic, dietary, and behavioral factors than by environmental exposure appeared to drive urinary levels. These findings suggest that urinary levoglucosan is not a suitable biomarker for assessing residential wood smoke exposure, with similar conclusions drawn for mannosan and galactosan. Full article

This paper investigates federated learning (FL) for cross-site scripting (XSS) detection under out-of-distribution (OOD) drift. Real-world XSS traffic involves fragmented attacks, heterogeneous benign inputs, and client imbalance, which erode conventional detectors. To simulate this, we construct two structurally divergent datasets: one with obfuscated, mixed-structure samples and another with syntactically regular examples, inducing structural OOD in both classes. We evaluate GloVe, GraphCodeBERT, and CodeT5 in both centralised and federated settings, tracking embedding drift and client variance. FL consistently improves OOD robustness by averaging decision boundaries from cleaner clients. Under FL scenarios, CodeT5 achieves the best aggregated performance (97.6% accuracy, 3.5% FPR), followed by GraphCodeBERT (96.8%, 4.7%), but is more stable on convergence. GloVe reaches a competitive final accuracy (96.2%) but exhibits a high instability across rounds, with a higher false positive rate (5.5%) and pronounced variance under FedProx. These results highlight the value and limits of structure-aware embeddings and support FL as a practical, privacy-preserving defence within OOD XSS scenarios. Full article

In scientific analyses, measurement errors in data can significantly impact statistical inferences, and ignoring them may lead to biased and invalid results. This study focuses on the estimation of the residual extropy function, in the presence of measurement errors. We developed an estimator for the extropy function and established its asymptotic properties. A comprehensive simulation study evaluates the performance of the proposed estimators under various error scenarios, while their practical utility and precision are demonstrated through an application to a real-world data set. Full article