Search Results (7,500)

This study explores the potential for afforestation in Portugal that could balance wood and non-wood forest production under future climate change scenarios. The Climate Envelope Models (CEM) approach was employed with three main objectives: (1) to model the current distribution of key Portuguese forest species—eucalypts, maritime pine, umbrella pine, chestnut, and cork oak—based on their suitability for wood and non-wood production; (2) to project their potential distribution for the years 2070 and 2090 under two Shared Socioeconomic Pathway (SSP) scenarios: SSP2–4.5 (moderate) and SSP5–8.5 (high emissions); and (3) to generate integrated species distribution maps identifying both current and future high-suitability zones to support afforestation planning, reflecting climatic compatibility under fixed thresholds. Species’ current CMEs were produced using an additive Boolean model with a set of environmental variables (e.g., temperature-related and precipitation-related, elevation, and soil) specific to each species. Species’ current CEMs were validated using forest inventory data and the official Land Use and Land Cover (LULC) map of Portugal, and a good agreement was obtained (>99%). By the end of the 21st century, marked reductions in species suitability are projected, especially for chestnut (36%–44%) and maritime pine (25%–35%). Incorporating future suitability projections and preventive silvicultural practices into afforestation planning is therefore essential to ensure climate-resilient and ecologically friendly forest management. Full article

This paper presents a closed-loop price–dispatch framework for park-scale virtual power plants (VPPs) with coupled electric–thermal processes under high penetrations of photovoltaics (PVs) and electric vehicles (EVs). The outer layer clears time-varying prices for operator electricity, operator heat, and user feed-in using an improved particle swarm optimizer with adaptive coefficients and velocity clamping. Given these prices, the inner layer executes a lightweight linear source decomposition with feasibility projection that enforces transformer limits, combined heat-and-power (CHP) and boiler constraints, ramping, energy balances, and EV state-of-charge requirements. PV uncertainty is represented by a small set of scenarios and a conditional value-at-risk (CVaR) term augments the welfare objective to control tail risk. On a typical winter day case, the coordinated setting aligns EV charging with solar hours, reduces evening grid imports, and improves a social welfare proxy while maintaining interpretable price signals. Measured outcomes include 99.17% PV utilization (95.14% self-consumption and 4.03% routed to EV charging) and a reduction in EV charging cost from CNY 304.18 to CNY 249.87 (−17.9%) compared with an all-from-operator benchmark; all transformer, CHP/boiler, and EV constraints are satisfied. The price loop converges within several dozen iterations without oscillation. Sensitivity studies show that increasing risk weight lowers CVaR with modest welfare trade-offs, while wider price bounds and higher EV availability raise welfare until physical limits bind. The results demonstrate an effective, interpretable, and reproducible pathway to integrate market signals with engineering constraints in park VPP operations. Full article

Manual pill identification is often inefficient and error-prone due to the large variety of medications and frequent visual similarity among pills, leading to misuse or dispensing errors. These challenges are exacerbated when pill imprints are engraved, curved, or irregularly arranged, conditions under which conventional optical character recognition (OCR)-based methods degrade significantly. To address this problem, we propose GO-PILL, a geometry-aware OCR pipeline for robust pill imprint recognition. The framework extracts text centerlines and imprint regions using the TextSnake algorithm. During imprint refinement, background noise is suppressed and contrast is enhanced to improve the visibility of embossed and debossed imprints. The imprint localization and alignment stage then rectifies curved or obliquely oriented text into a linear representation, producing geometrically normalized inputs suitable for OCR decoding. The refined imprints are processed by a multimodal OCR module that integrates a non-autoregressive language–vision fusion architecture for accurate character-level recognition. Experiments on a pill image dataset from the U.S. National Library of Medicine show that GO-PILL achieves an F1-score of 81.83% under set-based evaluation and a Top-10 pill identification accuracy of 76.52% in a simulated clinical scenario. GO-PILL consistently outperforms existing methods under challenging imprint conditions, demonstrating strong robustness and practical feasibility. Full article

In order to solve the problems of small-target detection in UAV aerial photography, such as small scale, blurred features and complex background interference, this article proposes the ACS-YOLOv8s method to optimize the YOLOv8s network: notably, most small man-made targets in UAV aerial scenes (e.g., small vehicles, micro-drones) inherently possess symmetry, a key geometric attribute that can significantly enhance the discriminability of blurred or incomplete target features, and thus symmetry-aware mechanisms are integrated into the aforementioned improved modules to further boost detection performance. The backbone network introduces an adaptive feature enhancement module, the edge and detail representation of small targets is enhanced by dynamically modulating the receptive field with deformable attention while also capturing symmetric contour features to strengthen the perception of target geometric structures; a cascaded multi-receptive field module is embedded at the end of the trunk to integrate multi-scale features in a hierarchical manner to take into account both expressive ability and computational efficiency with a focus on fusing symmetric multi-scale features to optimize feature representation; the neck is integrated with a spatially adaptive feature modulation network to achieve dynamic weighting of cross-layer features and detail fidelity and, meanwhile, models symmetric feature dependencies across channels to reduce the loss of discriminative information. Experimental results based on the VisDrone2019 data set show that ACS-YOLOv8s is superior to the baseline model in precision, recall, and mAP indicators, with mAP50 increased by 2.8% to 41.6% and mAP50:90 increased by 1.9% to 25.0%, verifying its effectiveness and robustness in small-target detection in complex drone aerial-photography scenarios. Full article

The exponential growth of data generated by modern digital applications, including systems inspired by Industrial Internet of Things (IIoT) requirements, has accelerated the adoption of NoSQL databases due to their scalability, flexibility, and performance advantages over traditional relational systems. Among document-oriented solutions, MongoDB and RavenDB stand out due to their architectural features and their ability to manage dynamic, large-scale datasets. This paper presents a comparative analysis of MongoDB and RavenDB, focusing on the performance of fundamental CRUD (Create, Read, Update, Delete) operations. To ensure a controlled performance evaluation, a mobile and web application for managing product orders was implemented as a case study inspired by IIoT data characteristics, such as high data volume and frequent transactional operations, with experiments conducted on datasets ranging from 1000 to 1,000,000 records. Beyond the core CRUD evaluation, the study also investigates advanced operational scenarios, including joint processing strategies (lookup versus document inclusion), bulk data ingestion techniques, aggregation performance, and full-text search capabilities. These complementary tests provide deeper insight into the systems’ architectural strengths and their behavior under more complex and data-intensive workloads. The experimental results highlight MongoDB’s consistent performance advantage in terms of response time, particularly with large data volumes, while RavenDB demonstrates competitive behavior and offers additional benefits such as built-in ACID compliance, automatic indexing, and optimized mechanisms for relational retrieval and bulk ingestion. The analysis does not propose a new benchmarking methodology but provides practical insights for selecting an appropriate document-oriented database for data intensive mobile and web application contexts, including IIoT-inspired data characteristics, based on a controlled single-node experimental setting, while acknowledging the limitations of a single-host experimental environment. Full article

Although greenhouse gas emissions have significantly reduced, the European Union still faces a major challenge in meeting its 2050 net-zero goal set under the European Green Deal. Focusing on the impacts of population, economic output, and carbon intensity of economy, this study employs Index Decomposition Analysis to estimate the reductions in carbon intensity needed to reach this target. The findings show that the extent of the technical effort required for decarbonization is much influenced by economic expansion. Under a 3% annual Gross Domestic Product growth scenario, the EU’s carbon intensity of economy must decline by 11.8% per year, which is a particularly demanding rate given the already low baseline. The decomposition also quantifies the technological challenge: under high growth, up to 5867 MtCO₂ in reductions would be needed by 2050 (compared with 1990), with Carbon Capture and Storage (CCS) contributing only 10–15%. In contrast, in zero- or negative-growth scenarios, required reductions fall to 4923–4594 MtCO₂, with CCS accounting for up to 50–90%. These results show that decarbonization in EU industrial sectors requires systemic transformations and strategic CCS deployment. A balanced approach, limiting economic growth and increasing innovation, appears essential to achieve the climate neutrality target. Full article

The rapid advancement of communication systems has heightened the demand for efficient and robust modulation recognition. Conventional deep learning-based methods, however, often struggle in practical few-shot scenarios where acquiring sufficient labeled training data is prohibitive. To bridge this gap, this paper proposes a hybrid transfer learning (HTL) approach that synergistically combines the representation power of deep feature extraction with the flexibility and stability of traditional machine learning (ML) classifiers. The proposed method capitalizes on knowledge transferred from large-scale auxiliary datasets through pre-training, followed by few-shot adaptation using simple ML classifiers. Multiple classical ML classifiers are incorporated and evaluated within the HTL framework for few-shot modulation recognition (FSMR). Comprehensive experiments demonstrate that HTL consistently outperforms existing baseline methods in such data-scarce settings. Furthermore, a detailed analysis of several key parameters is conducted to assess their impact on performance and to inform deployment in practical environments. Notably, the results indicate that the K-nearest neighbor classifier, owing to its instance-based and non-parametric nature, delivers the most robust and generalizable performance within the HTL paradigm, offering a promising solution for reliable FSMR in real-world applications. Full article

Effective disaster management is critical for safeguarding lives, infrastructure and economies in an era of escalating natural hazards like floods and landslides. Despite advanced early-warning systems and coordination frameworks, a persistent “last-mile” challenge undermines response effectiveness: transforming fragmented and unstructured multimodal data into timely and accountable field actions. This paper introduces ResQConnect, a human-centered, AI-powered multimodal multi-agent platform that bridges this gap by directly linking incident intake to coordinated disaster response operations in hazard-prone regions. ResQConnect integrates three key components. It uses an agentic Retrieval-Augmented Generation (RAG) workflow in which specialized language-model agents extract metadata, refine queries, check contextual adequacy and generate actionable task plans using a curated, hazard-specific knowledge base. The contribution lies in structuring the RAG for correctness, safety and procedural grounding in high-risk settings. The platform introduces an Adaptive Event-Triggered (AET) multi-commodity routing algorithm that decides when to re-optimize routes, balancing responsiveness, computational cost and route stability under dynamic disaster conditions. Finally, ResQConnect deploys a compressed, domain-specific language model on mobile devices to provide policy-aligned guidance when cloud connectivity is limited or unavailable. Across realistic flood and landslide scenarios, ResQConnect improved overall task quality scores from 61.4 to 82.9 (+21.5 points) over a standard RAG baseline, reduced solver calls by up to 85% compared to continuous re-optimization while remaining within 7–12% of optimal response time, and delivered fully offline mobile guidance with sub-500ms response latency and 54 tokens/s throughput on commodity smartphones. Overall, ResQConnect demonstrates a practical and resilient approach to AI-augmented disaster response. From a sustainability perspective, the proposed system contributes to Sustainable Development Goal (SDG) 11 by improving the speed and coordination of disaster response. It also supports SDG 13 by strengthening adaptation and readiness for climate-driven hazards. ResQConnect is validated using real-world flood and landslide disaster datasets, ensuring realistic incidents, constraints and operational conditions. Full article

Self-cleansing intrusion-tolerant systems mitigate attacker intrusions and control through periodic recovery, thereby enhancing both availability and security. However, vulnerabilities in the control link render these systems susceptible to request forgery attacks. Furthermore, existing research on the modeling and performance analysis of such systems remains insufficient. To address these issues, this paper introduces an authentication mechanism to fortify control link security and employs Generalized Stochastic Petri Nets for system evaluation. We constructed Petri net models for three distinct scenarios: a traditional system, a system compromised by forged controller requests, and a system fortified with authentication mechanism. Subsequently, isomorphic Continuous-Time Markov Chains were derived to facilitate theoretical analysis. Quantitative evaluations were performed by deriving steady-state probabilities and conducting simulations on the PIPE platform. To further assess practicality, we conduct scalability analysis under varying system scales and parameter settings, and implement a prototype in a virtualized testbed to experimentally validate the analytical findings. Evaluation results indicate that authentication mechanism ensures the reliable execution of cleansing strategies, thereby improving system availability, enhancing security, and mitigating data leakage risks. Full article

The particle exhaust system plays a pivotal role in fusion reactors and is essential for ensuring both the feasibility and sustained operation of the fusion reaction. For the successful development of such a system, density control is of great importance and some key design parameters include the neutral gas pressure and the resulting particle fluxes. This study presents a simplified conductance-based model for estimating neutral gas pressure distributions in the particle exhaust system of fusion reactors, focusing specifically on the sub-divertor region. In the proposed model, the pumping region is represented as an interconnected set of reservoirs and channels. Mass conservation and conductance relations, appropriate for all flow regimes, are applied. The model was benchmarked against complex 3D DIVGAS simulations across representative operating scenarios of the Wendelstein 7-X (W7-X) stellarator. Despite geometric simplifications, the model is capable of predicting pressure values at several key locations inside the particle exhaust area of W7-X, as well as various types of particle fluxes. The developed model is computationally efficient for large-scale parametric studies, exhibiting an average deviation of approximately 20%, which indicates reasonable predictive accuracy considering the model simplifications and the flow problem complexity. Its application may assist early-stage engineering design, pumping performance improvement, and operational planning for W7-X and other future fusion reactors. Full article

Human Activity Recognition (HAR) has become crucial across various domains, including healthcare, smart homes, and security systems, owing to the proliferation of Internet of Things (IoT) devices. Several Machine Learning (ML) techniques, including Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM), have been proposed for HAR. However, they are still deficient in addressing the challenges of noisy features and insufficient data. This paper introduces a novel approach to tackle these two challenges, employing a Deep Learning (DL) Ensemble-Based Stacking Neural Network (SNN) combined with Generative Adversarial Networks (GANs) for HAR based on ambient sensors. Our proposed deep learning ensemble-based approach outperforms traditional ML techniques and enables robust and reliable recognition of activities in real-world scenarios. Comprehensive experiments conducted on six benchmark datasets from the CASAS smart home project demonstrate that the proposed stacking framework achieves superior accuracy on five out of six datasets when compared to literature-reported state-of-the-art baselines, with improvements ranging from 3.36 to 39.21 percentage points and an average gain of 13.28 percentage points. Although the baseline marginally outperforms the proposed models on one dataset (Aruba) in terms of accuracy, this exception does not alter the overall trend of consistent performance gains across diverse environments. Statistical significance of these improvements is further confirmed using the Wilcoxon signed-rank test. Moreover, the ASGAN-augmented models consistently improve macro-F1 performance over the corresponding baselines on five out of six datasets, while achieving comparable performance on the Milan dataset. The proposed GAN-based method further improves the activity recognition accuracy by a maximum of 4.77 percentage points, and an average of 1.28 percentage points compared to baseline models. By combining ensemble-based DL with GAN-generated synthetic data, a more robust and effective solution for ambient HAR addressing both accuracy and data imbalance challenges in real-world smart home settings is achieved. Full article

Offensive-language detection systems that perform well at a given point in time often degrade as linguistic patterns evolve, particularly in dialectal Arabic social media, where new terms emerge and familiar expressions shift in meaning. This study investigates temporal linguistic drift in Saudi-dialect offensive-language detection through a systematic evaluation of continual-learning approaches. Building on the Saudi Offensive Dialect (SOD) dataset, we designed test scenarios incorporating newly introduced offensive terms, context-shifting expressions, and varying proportions of historical data to assess both adaptation and knowledge retention. Eight continual-learning configurations—Experience Replay (ER), Elastic Weight Consolidation (EWC), Low-Rank Adaptation (LoRA), and their combinations—were evaluated across five test scenarios. Results show that models without continual-learning experience a 13.4-percentage-point decline in F1-macro on evolved patterns. In our experiments, Experience Replay achieved a relatively favorable balance, maintaining 0.812 F1-macro on historical data and 0.976 on contemporary patterns (KR = −0.035; AG = +0.264), though with increased memory and training time. EWC showed moderate retention (KR = −0.052) with comparable adaptation (AG = +0.255). On the SimuReal test set—designed with realistic class imbalance and only 5% drift terms—ER achieved 0.842 and EWC achieved 0.833, compared to the original model’s 0.817, representing modest improvements under realistic conditions. LoRA-based methods showed lower adaptation in our experiments, likely reflecting the specific LoRA configuration used in this study. Further investigation with alternative settings is warranted. Full article

Multi-UAV (unmanned aerial vehicle) target enclosing control is one of the key technologies for achieving cooperative tasks. It faces limitations in communication resources and task framework separation. To address this, a distributed cooperative control strategy is proposed based on dynamic time-varying formation description and event-triggering mechanism. Firstly, a formation description method based on a geometric transformation parameter set is established to uniformly describe the translation, rotation, and scaling movements of the formation, providing a foundation for time-varying formation control. Secondly, a cooperative architecture for adaptive target enclosing tasks is designed. This architecture achieves an organic combination of formation control and target enclosing in a unified framework, thereby meeting flexible transitions between multiple formation patterns such as equidistant surrounding and variable-distance enclosing. Thirdly, a distributed dynamic event-triggered cooperative enclosing controller is developed. This strategy achieves online adjustment of communication thresholds through internal dynamic variables, significantly reducing communication while strictly ensuring system performance. By constructing a Lyapunov function, the stability and Zeno free behavior of the closed-loop system are proven. The simulation results verify this strategy, showing that this strategy can significantly reduce communication frequency while ensuring enclosing accuracy and formation consistency and effectively adapt to uniform and maneuvering target scenarios. Full article

Monitoring muscle fatigue is essential to ensure safety and support activity in populations such as the elderly. This study introduces a novel deep learning framework for classifying muscle fatigue levels using data from wireless surface electromyographic sensors, with the long-term goal of supporting applications in Ambient Assisted Living. A new dataset was collected from healthy elderly and non-elderly adults performing dynamic tasks under controlled conditions, with muscle fatigue levels labelled through self-assessment. The proposed method employs a pipeline that transforms one-dimensional electromyographic signals into two-dimensional time–frequency images (scalograms) using the Continuous Wavelet Transform, which are then classified by a fine-tuned, pre-trained Convolutional Neural Network. These images are then classified by pretrained Convolutional Neural Networks on large-scale image datasets. The classification pipeline includes an initial binary discrimination between non-fatigued and fatigued conditions, followed by a refined three-level classification into No Fatigue, Moderate Fatigue, and Hard Fatigue. The system achieved an accuracy of 98.6% in the binary task and 95.6% in the multiclass setting. This integrated transfer learning pipeline outperformed traditional Machine Learning methods based on manually extracted features, which reached a maximum of 92% accuracy. These findings highlight the robustness and generalizability of the proposed approach, supporting its potential as a real-time, non-invasive muscle fatigue monitoring solution tailored to Ambient Assisted Living scenarios. Full article

Target-oriented grasping has become increasingly important in household and industrial environments, and deploying such systems on mobile robots is particularly challenging due to limited computational resources. To address these limitations, we present an efficient framework for real-time target-oriented grasping on resource-constrained platforms, supporting both click-based grasping for unknown objects and category-based grasping for known objects. To reduce model complexity while maintaining detection accuracy, YOLOv8 is compressed using a structured pruning method. For grasp pose generation, a pretrained GR-ConvNetv2 predicts candidate grasps, which are restricted to the target object using masks generated by MobileSAMv2. A geometry-based correction module then adjusts the position, angle, and width of the initial grasp poses to improve grasp accuracy. Finally, extensive experiments were carried out on the Cornell and Jacquard datasets, as well as in real-world single-object, cluttered, and stacked scenarios. The proposed framework achieves grasp success rates of 98.8% on the Cornell dataset and 95.8% on the Jacquard dataset, with over 90% success in real-world single-object and cluttered settings, while maintaining real-time performance of 67 ms and 75 ms per frame in the click-based and category-specified modes, respectively. These experiments demonstrate that the proposed framework achieves high grasping accuracy and robust performance, with a efficient design that enables deployment on mobile and resource-constrained robots. Full article