Search Results (1,168)

With the rapid development of Intelligent Transportation Systems (ITSs), traffic prediction, a crucial component of ITSs, has garnered growing scholarly attention. The appli-cation of deep learning into traffic prediction has emerged as a prominent research direction, especially amid the rapid advancement of pretrained large language models (LLMs), which offer substantial benefits in time-series analysis through cross-modal knowledge transfer. In response to this advancement, this study introduces an innovative model for traffic flow prediction, designated as WGLLM. To capture spatiotemporal characteristics inherent in traffic flow data, this model incorporates a sequence embedding layer constructed on the stationary wavelet transform (SWT) and long short-term memory (LSTM), in conjunction with a spatial embedding layer founded on graph convolutional networks (GCNs). Additionally, a fully connected layer is utilized to integrate embeddings into the LLMs for comprehensive global dependency analysis. To verify the effectiveness of the proposed approach, experiments were carried out on two real traffic flow datasets. The experimental results demonstrate that WGLLM achieves superior predictive performance compared to multiple mainstream baseline models, accompanied by a significant enhancement in prediction accuracy. Full article

The continuous current dissipation of direct current grounding electrodes generates intense Joule heat, causing severe soil moisture loss and localized thermal runaway. Traditional static models ignore the temperature-dependent nature of soil parameters, leading to dangerous underestimations of actual temperature rises and thermal risks. To address this critical issue, this study establishes a bidirectional dynamic electro-thermal coupled model for a vertical grounding electrode using COMSOL Multiphysics. Comparative analysis demonstrates that the dynamic model accurately reproduces the late-stage accelerated temperature rise observed in experiments, proving its necessity over static methods. Simulations reveal that increased soil resistivity governs heat generation and directly causes a dramatic surge in both grounding resistance and maximum step voltage. In two-layer heterogeneous soils, current is forced into lower-resistivity regions, triggering extreme localized overheating. To mitigate this, expanding the cross-sectional radius of the coke bed effectively suppresses the thermal concentration. These findings provide quantitative evidence and non-uniform design guidelines for the safe operation and thermal protection of grounding electrodes under complex geological conditions. Full article

Background: Blockchain is increasingly proposed to strengthen pharmaceutical traceability, anti-counterfeiting, and chain of custody in multi-actor supply chains, but the evidence base remains heterogeneous in technical rigor and operational clarity. Methods: We conducted a mapping review of Scopus and Web of Science to map publication patterns, identify dominant thematic configurations, and compare citation-salient studies across recurring solution profiles and operational design dimensions. The final corpus comprised 103 records. Results: The literature expanded rapidly from 2019 to 2025, with notable geographic concentration and dissemination mainly through technically focused outlets. Keyword analysis identified a core traceability theme, an implementation stream centered on smart contracts, Ethereum, and security, and additional streams involving vaccines and regulatory or credentialing concerns. Citation-salient studies clustered into implemented systems and prototypes, architecture or framework proposals, and contextual maturity or decision-layer evidence. Across these profiles, transferability depended less on platform choice than on governance and access-control assumptions, modular smart contract roles, and verifiable on-chain/off-chain data placement. Conclusions: Chain-of-custody semantics and evaluation methods remain inconsistently formalized, limiting cross-study comparability and the interpretability of operational claims. Benchmark-oriented assessments and minimal reporting standards specifying governance parameters, logistics scope and checkpoints, workload, measurement conditions, and concrete evidence artifacts are needed. Full article

In karst tunnel engineering, water-filled cavities located above the tunnel crown, under the combined effects of excavation disturbance and hydraulic pressure, are prone to triggering water and mud inrush disasters. The thickness of the water-resisting rock layer is therefore a key factor controlling the stability of the surrounding rock. To address the difficulty in accurately characterizing the mechanical behavior of the crown of horseshoe-shaped tunnels using conventional circular plate or beam models, this study innovatively develops an explicit analytical model for the minimum safe thickness of the water-resisting rock layer based on clamped elliptical thin plate theory and Kirchhoff plate theory, incorporating the influence of cross-sectional geometry. Parametric sensitivity analysis indicates that both karst water pressure and tunnel crown height significantly amplify the required minimum safe thickness, whereas an increase in the tensile strength of the surrounding rock effectively reduces the thickness demand. Specifically, when the karst water pressure increases from 2.5 MPa to 4.5 MPa, the minimum safe thickness rises from 7.5 m to 10.0 m, showing an approximately linear growth trend. The analytical model is further validated through numerical simulations under different “water pressure–thickness” conditions. The results demonstrate that at the calculated recommended thickness, the surrounding rock achieves stable convergence after excavation. High tensile stress and elevated pore pressure zones are mainly concentrated near the tunnel crown, without the formation of through-going tensile failure. Engineering application indicates that the proposed model can provide a quantitative basis for the design of water-resisting rock layer thickness and the assessment of water inrush risk in karst tunnels. Full article

This article presents a novel compact high-Q bandpass filter (BPF) utilizing a 3-D stacked stripline configuration. T-shaped stepped impedance resonators (SIRs) are employed to achieve miniaturization. By folding the filter geometry from an inline arrangement into a U-shape along the broadside direction, both broadside and edge coupling structures are realized, enabling various cross-coupling schemes for flexible placement of transmission zeros (TZs). A comprehensive analysis of both electric and magnetic coupling structures is conducted to support the overall filter design. To validate the concept, a tenth-order general Chebyshev BPF prototype centered at 3.485 GHz with a 1 dB bandwidth of 380 MHz is designed, fabricated, and measured. The filter is constructed by vertically soldering two patterned sheet metal layers together with three stacked cavities. Despite having an electrical size of only 0.58 × 0.23 × 0.19 λ_g³, the filter exhibits a high unloaded Q-factor (Q_u) of 1200, along with up to six TZs and a spurious-free frequency range extending to 12 GHz. Measured results show an insertion loss of 0.58 dB at the center frequency and a return loss of better than 20 dB within the passband, demonstrating favorable agreement with simulations. Featuring solid electrical performance, the proposed filter is ideally suited for 5G and 5G-Advanced (5G-A) communication base stations. Full article

Weight reduction and dynamic performance optimization are critical for airborne direct-drive VICTS satellite communication antennas, which require lightweight, high-speed, and high-precision rotation. Traditional vibration suppression methods, such as uniform support layout and added damping, rely heavily on empirical trial and error, lack targeted modal control, and cannot balance lightweight design with dynamic stiffness. To address these issues, this paper proposes a wave-theory-based dynamic modeling and rapid optimization method for multi-layer rotating components in direct-drive VICTS antennas. The kinematic model of the rotating ring and ball revolution excitation are derived using the annular wave equation and bearing kinematics. A Modal Blocking Mechanism is established: placing support balls at positions satisfying the half-wavelength constraint suppresses target mode shapes via wave interference, achieving vibration attenuation at the source. A homogenization equivalent method based on RVE is developed for irregular cross-section rings, yielding analytical expressions for in-plane equivalent elastic modulus and out-of-plane equivalent shear modulus. These parameters are integrated into the wave equation to analytically solve vibration modes, avoiding iterative finite element computations. A rapid multi-objective optimization framework is then constructed, minimizing the structural weight and maximizing the modal separation interval under dynamic stiffness and excitation frequency constraints. Numerical simulations, FE analysis, and prototype tests validate the method: the maximum analytical error is only 3.1%. Compared with uniform support designs, the optimized structure achieves a 40% weight reduction, a 40% increase in minimum modal separation, and a 65% reduction in the RMS tracking error. This work provides an efficient, deterministic dynamic design method for large-diameter ring structures, transforming vibration control from empirical adjustment into a precise, physics-informed optimization. Full article

Spatial qualities are central to architectural reasoning; yet, in studio-based education, they often remain implicit rather than structured as a shared analytical framework. This study examines how a multi-scalar taxonomy of spatial qualities can function as a collaborative analytical language in studio-based architectural education. Situated in Košanćićev venac and Dorćol, two historically layered areas of Belgrade’s old town, this study integrates expert spatial analysis with a student questionnaire administered across bachelor and master study levels. Empirical testing was conducted to evaluate structural coherence, conceptual differentiation and the distribution of spatial qualities across detail, architectural and urban drawing scales. The findings indicate consistent internal stability, clear differentiation among constructs and statistically significant cross-scale articulation. Form- and composition-related qualities showed high usability, while interpretative constructs were more variable. Master-level students demonstrated greater engagement with cognitive and interpretative constructs, indicating a shift toward more conceptually grounded design reasoning without affecting overall structural coherence. These results suggest that spatial qualities can operate as a level-independent analytical language, supporting inclusive participation, shared interpretation and structured dialogue within the design studio. By positioning spatial qualities as a collaborative pedagogical framework, this study contributes to interdisciplinary communication and more equitable engagement in architectural education. Full article

Amidst global urbanization and rising psychological stress, urban green spaces are increasingly recognized as critical infrastructure for sustainable urban development and public health. However, the mechanisms by which summer vegetation structure mediates both physiological and psychological restoration, and the interplay between these two dimensions, remain poorly understood. Understanding these mechanisms is essential for designing sustainable, health-promoting urban environments that can support growing urban populations in a warming climate. This study employed a controlled field experiment in Xi’an during summer to examine the effects of five vegetation structure types (Single-Layer Grassland, single-layer woodland, tree–shrub–grass composite woodland, tree–grass composite woodland, and a non-vegetated square) on university students’ physiological (heart rate variability) and psychological (perceived restorativeness and affective states) restoration. Following stress induction, 300 participants engaged with the green spaces through both quiet sitting and walking. The results revealed three key findings: (1) the tree–shrub–grass composite woodland consistently showed the most favorable trends other vegetation types across all psychological restoration dimensions, while also showing favorable trends in physiological recovery, underscoring the importance of structural complexity for restorative quality; (2) walking significantly enhanced physiological recovery compared to seated observation across all settings, confirming the role of physical activity as a critical activator of green space benefits; (3) correlation analysis identified a specific cross-system association: the R-R interval recovery value showed a weak but significant correlation with positive affect (PA) scores, suggesting that physiological calmness and positive emotional experience are linked, yet their weak coupling under short-term exposure indicates they may operate as parallel processes with distinct temporal dynamics. These findings indicate that the restorative potential of summer green spaces emerges from an integrated framework combining vegetation complexity and activity support. We propose that future sustainable landscape design should prioritize multi-layered vegetation structures as nature-based solutions that simultaneously enhance human well-being and urban resilience. These findings provide empirical evidence for integrating health-promoting green infrastructure into sustainable urban planning frameworks, supporting multiple Sustainable Development Goals (SDGs), including SDG 3 (Good Health and Well-being), SDG 11 (Sustainable Cities and Communities), and SDG 13 (Climate Action). Full article

To investigate the dominant factors and instability mechanism of surrounding rock deformation in cross-mining roadways, a systematic study was conducted using theoretical analysis, numerical simulation, and response surface methodology to examine the influence of various factors on surrounding rock stability. First, the theoretical model was refined by introducing a lithology coefficient of the load-transfer layer, thereby improving its engineering applicability. Subsequently, numerical simulations and response surface experiments were employed to analyze the effects of key factors, including the vertical distance between the working face and the roadway, the horizontal distance between the working face and the roadway, the burial depth of the roadway, the mining height of the working face, and the lithology of the load-transfer layer. The analysis results indicate that the vertical distance, horizontal distance, and lithology of the load-transfer layer are negatively correlated with roadway roof displacement, whereas the burial depth and mining height are positively correlated. The p-values for all factors were less than 0.0001. The order of significance of the influencing factors is as follows: vertical distance > horizontal distance > burial depth > mining height > lithology of the load-transfer layer. Among these, the vertical distance has the most significant effect on roadway deformation and exhibits notable interaction effects with burial depth and horizontal distance. Based on these findings, given that construction conditions cannot be altered, modifying the lithology of the load-transfer layer was selected as the control measure. Directional long-hole hydraulic fracturing for roof cutting and pressure relief was implemented in the roof of the return airway in the No. 6 mining district. Field monitoring results show that hydraulic fracturing effectively interrupted the stress transmission path induced by mining activities, transferring roof pressure to deeper strata. Consequently, the deformation of the surrounding rock was significantly reduced, the dynamic pressure effect was markedly alleviated, and the stability of the roadway was effectively controlled. The research results provide a theoretical basis for the design and control of cross-mining roadways under similar engineering conditions. Full article

Proteomics represents a fundamental layer for understanding the molecular complexity of solid tumors by quantifying protein abundance and capturing proteoforms and post-translational modifications undetected in genomics or transcriptomics analyses. As mass spectrometry-based technologies and public proteomics repositories have expanded, opportunities for large-scale data reuse have grown accordingly. Nevertheless, data availability has not been translated into straightforward reuse: differences in experimental design, acquisition strategies, quantification workflows and metadata quality still limit the reproducibility and cross-study comparability. In this review, proteomics data reuse is defined as the systematic reanalysis and integration of publicly available datasets to support precision oncology applications such as biomarker assessment and antibody–drug conjugate target prioritization. We discuss reuse as an end-to-end analytical process, focusing on data analysis workflows, harmonization strategies, and the impact of heterogeneous experimental and analytical choices on interoperability. The increased application of artificial intelligence in proteomics data integration and reuse is also addressed, highlighting its analytical potential while underscoring the risks of overinterpretation when biological context and data structure are not adequately considered. Using colorectal and prostate cancer as representative examples, we illustrate how proteomics data reuse can support biological discovery and translational research, while critically examining the factors that limit robustness and clinical relevance. Full article

Professional education increasingly requires graduates to make decisions in complex systems marked by multiple stakeholders, feedback, delays, uncertainty, and unintended consequences, yet systems thinking is still often taught as a set of disconnected tools rather than as an integrated professional practice. This conceptual paper adopts an integrative theory-building approach to develop a unified architecture for systems thinking in professional education, drawing purposively on systems traditions, practice-based learning, assessment scholarship, and emerging work on generative artificial intelligence (GenAI). The paper proposes four iterative practices (sensemaking and boundary setting, co-modelling and causal representation, intervention reasoning, and meta-learning) as the core architecture for learning systems thinking in professional contexts. It further translates this architecture into indicative implications for curriculum sequencing, authentic tasks, and assessment, while positioning GenAI as a cross-cutting support/risk layer that can assist iteration and critique but also introduce predictable risks such as fabricated causal links, overreliance, and false mastery. To address these risks, the paper outlines governance conditions based on traceability, uncertainty checks, stakeholder validation, and process-based assessment. Overall, the framework provides a design-oriented basis for teaching, assessing, and governing systems thinking in contemporary professional education and a foundation for future empirical testing. Full article

Vehicular infrared object detection is a key technology supporting autonomous driving systems to achieve all-weather environmental perception. However, infrared images inherently lack texture, resulting in blurred object contours. Additionally, deep network propagation severely erodes and loses feature information of distant tiny objects. To address the above issues, this study proposes a Full-Scale Feature Synergistic Perception Architecture for vehicular infrared object detection. This architecture first designs a Gradient-Informed Attention module, which initializes convolution kernels through physical gradient operators to inject geometric prior information into the network, enhancing the model’s perception capability of blurred object boundaries. Secondly, it constructs a Full-Scale Feature Pyramid containing a $P_2$ high-resolution feature layer to effectively recover the geometric detail features of distant tiny objects. Finally, it proposes a Scale-Aware Shared Head, which relies on a cross-scale parameter sharing mechanism to achieve extreme parameter compression, and simultaneously introduces deep semantic information to form strong constraints, suppressing noise interference in shallow features. Experimental results on the FLIR v2 and M3FD datasets show that the proposed architecture exhibits excellent detection performance. On FLIR v2, it raises mAP@50 to 64.06% (6.51% relative gain vs. YOLOv11) while maintaining 547 FPS inference speed, achieving an optimal accuracy–efficiency balance. Full article

The Internet of Things (IoT) plays an increasingly central role in healthcare by enabling continuous patient monitoring, remote diagnosis, and data-driven clinical decision-making through interconnected medical devices and sensing infrastructures. Despite these advances, IoT healthcare systems remain constrained by persistent challenges related to data privacy, computational efficiency, scalability, and regulatory compliance. Federated learning (FL) reduces reliance on centralised data aggregation but remains vulnerable to inference-based privacy risks, while edge-oriented approaches are limited by device heterogeneity and restricted computational and energy resources; the deployment of large language models (LLMs) further exacerbates concerns surrounding privacy exposure, communication overhead, and practical feasibility. This study introduces Trustworthy Intelligence (TI) as a guiding framework for privacy-preserving distributed intelligence in IoT healthcare, explicitly integrating predictive performance, privacy protection, and deployment-oriented system design. Within this framework, split learning (SL) is examined as a core architectural mechanism and extended to support split-aware LLM integration across heterogeneous devices, supported by a structured taxonomy spanning architectural configurations, system adaptation strategies, and evaluation considerations. The study establishes a systematic mapping between SL design choices and representative healthcare scenarios, including wearable monitoring, multi-modal data fusion, clinical text analytics, and cross-institutional collaboration, and analyses key technical challenges such as activation-level privacy leakage, early-round vulnerability, reconstruction risks, and communication–computation trade-offs. An energy- and resource-aware adaptive cut layer selection strategy is outlined to support efficient deployment across devices with varying capabilities. A proof-of-concept experimental evaluation compares the proposed SL–LLM framework with centralised learning (CL), federated learning (FL), and conventional SL in terms of training latency, communication overhead, model accuracy, and privacy exposure under realistic IoT constraints, providing system-level evidence for the applicability of the TI framework in distributed healthcare environments and outlining directions for clinically viable and regulation-aligned IoT healthcare intelligence. Full article

To tackle the challenges of high computational expense, limited detection accuracy, and imbalanced detection performance across multi-scale targets in rice disease identification within complex natural environments, we propose the Rice Disease Deformable Detection Transformer (RDD-DETR). This model serves as a full-scale detection framework based on the Deformable Detection Transformer (Deformable DETR). The model introduces a Rectified Linear Unit (ReLU)-enhanced lightweight linear attention module, which uses differentiated position coding and ReLU kernel mapping to reduce computational complexity. A cross-layer dynamic fusion and inter-layer supervision module is designed to break the serial dependence in decoders and strengthen interlayer supervision, enabling the decoder to generate more accurate and robust target representations. Furthermore, we design an optimization mechanism for sub-scale positioning loss to substantially boost detection accuracy across all target scales. Experiments on our custom RiceLeafDisease-RSOD dataset demonstrate that RDD-DETR achieves an average precision (AP) at Intersection over Union (IoU) threshold 0.5:0.95 of 0.7363 across all categories, surpassing the baseline model by 6.09%. Notably, detection accuracy improves by 6.10% for small targets, 6.61% for medium targets, and 5.42% for large targets. Evaluated on the validation set (671 images with 2482 labeled bounding boxes), the model achieves an AP at IoU threshold 0.5 of 0.9684 while reducing computational cost by 37.41% (from 136.02 to 85.1 Giga Floating Point Operations, GFLOPs) compared to the original Deformable DETR. These results validate RDD-DETR as an effective solution for accurate and efficient real-time rice disease monitoring in complex field environments. Full article

Underwater object detection plays a crucial role in the domain of marine engineering. Due to blur, uneven illumination and noise in underwater images, generic object detectors often fail to accurately detect underwater targets. Existing underwater object detection methods generally neglect the enhancement and refinement of multi-scale features, limiting further improvements in detection accuracy. In response to these challenges, we propose the Feature Enhancement and Alignment Pyramid Network (FEAPN), a novel underwater object detection framework. FEAPN consists of two key innovations. First, the Adaptive Feature Refinement Module (AFRM) is developed to adaptively enhance contextual features from complex backgrounds. Second, the Dual-path Feature Alignment Module (DFAM) is designed to align multi-scale features, utilizing cross-layer information to optimize feature representation. Extensive experiments demonstrate that FEAPN achieves state-of-the-art performance. Specifically, FEAPN achieves a 2.4% mAP improvement over the baseline and outperforms the current leading underwater detector by 1.2% mAP. Furthermore, the effectiveness of each component is validated through ablation studies. Full article