Search Results (1,021)

Under the dual pressures of climate change and rapid urbanization, extreme heat events pose growing risks to densely populated megaregions. The Guangdong–Hong Kong–Macao Greater Bay Area (GBA), a densely populated and economically vital region, serves as a critical hotspot for heat risk aggregation. This study develops a high-resolution multi-dimensional framework to assess the spatiotemporal evolution of its heat risk profile from 2000 to 2020. A Heat Risk Index (HRI) integrating heat hazard and vulnerability components to measure potential heat-related impacts is calculated as the product of the Heat Hazard Index (HHI) and Heat Vulnerability Index (HVI) for 1 km grids in GBA. The HHI integrates the frequency of hot days and hot nights. HVI incorporates population density, GDP, remote-sensing nighttime light data, and MODIS-based landscape indicators (e.g., NDVI, NDWI, and NDBI), with weights determined objectively using the static Entropy Weight Method to ensure spatiotemporal comparability. The findings reveal an escalation of heat risk, expanding at an average rate of 342 km² per year (p = 0.008), with the proportion of areas classified as high-risk or above increasing from 21.8% in 2000 to 33.3% in 2020. This trend was characterized by (a) a pronounced asymmetric warming pattern, with nighttime temperatures rising more rapidly than daytime temperatures; (b) high vulnerability dominated by the concentration of population and economic assets, as indicated by high EWM-based weights; and (c) isolated high-risk hotspots (Guangzhou and Hong Kong) in 2000, which have expanded into a high-risk belt across the Pearl River Delta’s industrial heartland, like Foshan seeing their high-risk area expand from 3.4% to 27.0%. By combining remote sensing and socioeconomic data, this study provides a transferable framework that moves beyond coarse-scale assessments to identify specific intra-regional risk hotspots. The resulting high-resolution risk maps offer a quantitative foundation for developing spatially explicit climate adaptation strategies in the GBA and other rapidly urbanizing megaregions. Full article

Here we present a high-resolution landslide susceptibility model for Guiyang County, China, developed to support sustainable disaster risk management. Our approach couples optimized positive and negative training samples with an ensemble of machine-learning algorithms to maximize predictive fidelity. We compiled a georeferenced inventory of 146 landslides by integrating historical records with systematic field validation. Sample optimization was central to our methodology: landslide presence points were refined via buffer-based dilution, and four classifiers—SVM, LDA, RF, and ET—were trained with identical covariate sets to ensure comparability. Three strategies for selecting pseudo-absences—buffering, low-slope filtering, and coupling with the IOE—were benchmarked. The Slope-IOE-O model, which synergizes low-gradient screening with entropy-weighted sampling, yielded the highest predictive capacity (AUC = 0.965). SHAP-based interpretability revealed that slope, monthly maximum rainfall, surface roughness, and elevation collectively dominate susceptibility, with pronounced non-linearities and interactions. Slope contribution peaks at 20–30°, monthly maximum rainfall exhibits a critical threshold near 225 mm, and the synergy between high roughness and road density amplifies landslide risk. Spatially, susceptibility follows a pronounced north–south gradient, with high-hazard corridors aligned along northern and southern mountain belts and the urban core of southern Guiyang County. By integrating rigorously curated training data with robust machine-learning workflows, this study provides a transferable framework for proactive landslide risk assessment, offering scientific support for sustainable land-use planning and resilient development in mountainous regions. Full article

Human Activity Recognition (HAR) is a key technology with applications in healthcare, security, smart environments, and sports analytics. Despite advances in deep learning, challenges remain in building models that are both efficient and generalizable due to the large scale and variability of video data. To address these issues, we propose a novel Dual-Teacher Knowledge Distillation (DTKD) framework tailored for HAR. The framework introduces three main contributions. First, we define static and dynamic activity classes in an objective and reproducible manner using optical-flow-based indicators, establishing a quantitative classification scheme based on motion characteristics. Second, we develop subset-specialized teacher models and design a hybrid scoring mechanism that combines teacher confidence with cross-entropy loss. This enables dynamic weighting of teacher contributions, allowing the student to adaptively balance knowledge transfer across heterogeneous activities. Third, we provide a comprehensive evaluation on the UCF101 and HMDB51 benchmarks. Experimental results show that DTKD consistently outperforms baseline models and achieves balanced improvements across both static and dynamic subsets. These findings validate the effectiveness of combining subset-aware teacher specialization with hybrid scoring. The proposed approach improves recognition accuracy and robustness, offering practical value for real-world HAR applications such as driver monitoring, healthcare, and surveillance. Full article

Aiming at the problems of high complexity in source domain data, inaccessibility of target domain data, and unknown fault patterns in real-world industrial scenarios for mechanical fault diagnosis, this paper proposes a Feature Decomposition-based Source-Free Universal Domain Adaptation (FD-SFUniDA) framework for mechanical equipment fault diagnosis. First, the CBAM attention module is incorporated to enhance the ResNet-50 convolutional network for extracting feature information from source domain data. During the target domain adaptation phase, singular value decomposition is applied to the weights of the pre-trained model’s classification layer, orthogonally decoupling the feature space into a source-known subspace and a target-private subspace. Then, based on the magnitude of feature projections, a dynamic decision boundary is constructed and combined with an entropy threshold mechanism to accurately distinguish between known and unknown class samples. Furthermore, intra-class feature consistency is strengthened through neighborhood-expanded contrastive learning, and semantic weight calibration is employed to reconstruct the feature space, thereby suppressing the negative transfer effect. Finally, extensive experiments under multiple operating conditions on rolling bearing and reciprocating mechanism datasets demonstrate that the proposed method excels in addressing source-free fault diagnosis problems for mechanical equipment and shows promising potential for practical engineering applications in fault classification tasks. Full article

In tightly coupled Industrial Control Systems (ICS), abnormal disturbances often propagate throughout the process, triggering a large number of time-correlated alarms that exceed the handling capacity of the operator. Consequently, a key challenge is how to leverage the directional and temporal characteristics of disturbance propagation to alleviate alarm overload. This paper proposes a delay-sensitive causal inference approach for industrial alarm analysis to address this problem. On the one hand, time delay estimation is introduced to precisely align the responses of two sensor sequences to disturbances, thereby improving the accuracy of causal relationship identification in the temporal domain. On the other hand, a multi-scale subgraph fusion strategy is designed to address the inconsistency in causal strength caused by disturbances of varying intensities. By integrating significant causal subgraphs from multiple scenarios into a unified graph, the method reveals the overall causal structure among alarm variables and provides guidance for alarm mitigation. To validate the proposed method, a case study is conducted on the Tennessee Eastman Process. The results demonstrate that the approach identifies causal relationships more accurately and reasonably and can effectively reduce the number of alarms by up to 51.6%. Full article

A novel pumped storage system using centrifugal pumps to transfer water between reservoirs in coastal hydropower plants has significantly mitigated grid instability. However, frequent start–stop operations of large vertical centrifugal pumps, which serve as the core equipment, severely affect the operational stability of these systems. In this study, the intrinsic connection between the cavitating flow field and irreversible losses during the process was analyzed using the entropy production theory. The time–frequency characteristics of pressure pulsation in pump were analyzed by using the continuous wavelet transform. The results indicate that with the reduction in the flow rate and rotational speed, the sheet cavitation at the impeller inlet rapidly weakens until it vanishes. The cavity cavitation within the draft tube commences to emerge in the turbine mode. Separation vortices are formed due to the mismatch in the flow angle at the impeller outlet. These vortices induce local cavitation, causing both a rapid energy loss increase and high-amplitude, low-frequency pressure pulsations. During transient processes, flow instabilities induce high-amplitude, low-frequency pressure pulsations within the stay vane region, with maximum amplitude attained during runaway condition. The research results provide a theoretical foundation for the stable operation of centrifugal pumps. Full article

We present a comprehensive information-theoretic evaluation of three widely used rigid water models (TIP3P, SPC, and SPC/$\epsilon$) through systematic analysis of water clusters ranging from single molecules to 11-molecule aggregates. Five fundamental descriptors—Shannon entropy, Fisher information, disequilibrium, LMC complexity, and Fisher–Shannon complexity—were calculated in both position and momentum spaces to quantify electronic delocalizability, localization, uniformity, and structural sophistication. Clusters containing 1, 3, 5, 7, 9, and 11 molecules (denoted 1 M, 3 M, 5 M, 7 M, 9 M, and 11 M) were selected to balance computational tractability with representative scaling behavior. Molecular dynamics simulations validated the force fields against experimental bulk properties (density, dielectric constant, self-diffusion coefficient), while statistical analysis using Shapiro–Wilk normality tests and Student’s t-tests ensured robust discrimination between models. Our results reveal distinct scaling behaviors that correlate with experimental accuracy: SPC/$\epsilon$ demonstrates superior electronic structure representation with optimal entropy–information balance and enhanced complexity measures, while TIP3P shows excessive localization and reduced complexity that worsen with increasing cluster size. The transferability from clusters to bulk properties is established through systematic convergence of information-theoretic measures toward bulk-like behavior. The methodology establishes information-theoretic analysis as a useful tool for comprehensive force field evaluation. Full article

This work investigates the influence of chemical composition, grain boundary (GB) type, and atomic distribution on the thermal conductivity of Hf–Nb–Ta–Zr refractory high-entropy alloys (RHEAs) via atomistic simulations. Three compositions—equiatomic HfNbTaZr (M1), Hf₁₀Nb₄₀Ta₁₀Zr₄₀ (M2), and Hf₄₀Nb₁₀Ta₄₀Zr₁₀ (M3)—were studied in single-crystalline and bicrystalline models containing Σ3 or Σ5 GBs. The effect of chemical short-range order (SRO) and GB segregation was probed by comparing results for non-relaxed structures with those obtained for corresponding materials relaxed using combined Monte Carlo/molecular dynamics (MC/MD) simulation. Material relaxation is accompanied by the formation of coherent nanoclusters (NbTa in M1, Nb or Zr in M2, Hf or Ta in M3) and Hf/Zr segregation to GBs. In single crystals, SRO reduces thermal conductivity by up to ~2.7% (e.g., from 3.66 to 3.56 W/m·K in M1), which is explained by the phonon scattering effect from matrix–cluster interfaces, densely distributed in the structures. In contrast, in certain bicrystals, the combined effects of GB healing and intragranular cluster coarsening lead to a 6.9% increase in thermal conductivity (from 4.59 to 4.93 W/m·K), despite the presence of high-energy Σ5 GBs. These results demonstrate that the interplay between SRO, GB segregation, and microstructural evolution governs phonon transport in RHEAs, revealing a counterintuitive pathway to enhance thermal conductivity through controlled atomic redistribution. Full article

Magnetohydrodynamic (MHD) flow and heat transfer in porous media are central to many engineering applications, including heat exchangers, MHD generators, and polymer processing. This study examines the boundary layer flow and thermal behavior of an electrically conducting viscous fluid over a porous stretching tube. The model accounts for nonlinear thermal radiation, internal heat generation/absorption, and Darcy–Forchheimer drag to capture porous medium resistance. Similarity transformations reduce the governing equations to a system of coupled nonlinear ordinary differential equations, which are solved numerically using the BVP4C technique with Response Surface Methodology (RSM) and sensitivity analysis. The effects of dimensionless parameters magnetic field strength (M), Reynolds number (Re), Darcy–Forchheimer parameter (Df), Brinkman number (Br), Prandtl number (Pr), nonlinear radiation parameter (Rd), wall-to-ambient temperature ratio (rw), and heat source/sink parameter (Q) are investigated. Results show that increasing M, Df, and Q suppresses velocity and enhances temperature due to Lorentz and porous drag effects. Higher Re raises pressure but reduces near-wall velocity, while rw, Rd, and internal heating intensify thermal layers. The entropy generation analysis highlights the competing roles of viscous, magnetic, and thermal irreversibility, while the Bejan number trends distinctly indicate which mechanism dominates under different parameter conditions. The RSM findings highlight that rw and Rd consistently reduce the Nusselt number (Nu), lowering thermal efficiency. These results provide practical guidance for optimizing energy efficiency and thermal management in MHD and porous media-based systems.: Full article

Background: This study investigated sex differences in foot arch structure and function, and their impact on postural control and energy flow during dynamic tasks. Findings aim to inform sex-specific training, movement assessment, and injury prevention strategies. Methods: A total of 108 participants (53 males and 55 females) underwent foot arch morphological assessments and performed a sit-to-stand (STS). Motion data were collected using an infrared motion capture system, three-dimensional force plates, and wireless surface electromyography. A rigid body model was constructed in Visual3D, and joint forces, segmental angular and linear velocities, center of pressure (COP), and center of mass (COM) were calculated using MATLAB. Segmental net energy was integrated to determine energy flow across different phases of the STS. Results: Arch stiffness was significantly higher in males. In terms of postural control, males exhibited significantly lower mediolateral COP frequency and anteroposterior COM peak velocity during the pre-seat-off phase, and lower COM displacement, peak velocity, and sample entropy during the post-seat-off phase compared to females. Conversely, males showed higher anteroposterior COM velocity before seat-off, and greater anteroposterior and vertical momentum after seat-off (p < 0.05). Regarding energy flow, males exhibited higher thigh muscle power, segmental net power during both phases, and greater shank joint power before seat-off. In contrast, females showed higher thigh joint power before seat-off and greater shank joint power after seat-off (p < 0.05). Conclusions: Significant sex differences in foot arch function influence postural control and energy transfer during STS. Compared to males, females rely on more frequent postural adjustments to compensate for lower arch stiffness, which may increase mechanical loading on the knee and ankle and elevate injury risk. Full article

Quantum-enhanced machine learning, encompassing both quantum algorithms and quantum-inspired classical methods such as tensor networks, offers promising tools for extracting structure from complex, high-dimensional data. In this work, we study the training dynamics of Matrix Product State (MPS) classifiers applied to three-class problems, using both fashion MNIST and hyperspectral satellite imagery as representative datasets. We investigate the phenomenon of grokking, where generalization emerges suddenly after memorization, by tracking entanglement entropy, local magnetization, and model performance across training sweeps. Additionally, we employ information-theory tools to gain deeper insights: transfer entropy is used to reveal causal dependencies between label-specific quantum masks, while O-information captures the shift from synergistic to redundant correlations among class outputs. Our results show that grokking in the fashion MNIST task coincides with a sharp entanglement transition and a peak in redundant information, whereas the overfitted hyperspectral model retains synergistic, disordered behavior. These findings highlight the relevance of high-order information dynamics in quantum-inspired learning and emphasize the distinct learning behaviors that emerge in multi-class classification, offering a principled framework to interpret generalization in quantum machine learning architectures. Full article

The heat generation of lithium-ion batteries is a critical parameter, as it significantly affects cell temperature. Poor thermal management can lead to elevated cell temperatures, accelerating side reactions, reducing cell lifetime, and, in extreme cases, causing thermal runaway. Therefore, understanding heat generation is crucial for the commercialization of emerging battery materials. Due to its high energy density, lithium–nickel–manganese–oxide (LNMO) is an attractive candidate for next-generation cathode materials; however, the composition of its heat generation is not yet fully understood. To address this, the state-of-charge (SoC)-dependent entropy coefficient and resistance of disordered LNMO cathodes are determined using the potentiometric method. The results show that both values are strongly influenced by the redox reactions of Ni and Mn. The entropy coefficient varies between 5.2 and −32.4 J ${\mathrm{mol}}^{-1}{\mathrm{K}}^{-1}$, depending on the SoC. Furthermore, the resistance exhibits a switching dependence on kinetics and mass transfer. The resulting heat flux calculations indicate that, at SoC < 20%, heat generation is dominated by the kinetic behavior of LNMO, leading to two exothermal peaks during discharge and one exothermal peak during charge. This behavior is validated through a comparison with a low-current calorimetric measurement. Full article

Studying the information transfer between the brain and muscles during archery can help us to understand the underlying mechanisms of corticomuscular coupling during motor learning. In this study, we recruited 26 novice archers as participants and calculated the transfer entropy (TE) between their EEG and EMG signals during the archery process. This was performed to assess the characteristics of corticomuscular coupling during archery and the impact of a period of archery training on this coupling. The results indicate that information transfer from EEG to EMG in the $\alpha$ and $\beta$ frequency bands predominates during archery, which may be related to the roles of $\alpha$ and $\beta$ frequency bands in inhibitory control and the sustained contraction of muscle stability. Additionally, the optimization of brain resource allocation resulting from a period of archery training is primarily reflected in the prefrontal cortex and motor cortex, where the information transfer from EEG to EMG decreases while activation related to inhibitory control increases. The intensity of corticomuscular coupling weakens with an increase in the number of arrows shot, but archery training reduces the impact of fatigue-induced changes on corticomuscular coupling. Full article

This work presents the development of a novel multi-scale modeling framework for investigating the beneficial impact of Ti-, Zr-, and V-doped magnesium hydride nanomaterials on hydrogen storage performance. The proposed model integrates atomistic-scale simulations based on density functional theory (DFT) with system-level dynamic heat and mass transfer modeling. At the nanoscale, DFT analysis provides key thermodynamic and kinetic parameters, including reaction enthalpy, entropy, and activation energy, which are incorporated into the macroscopic model to predict the hydrogenation behavior of MgH₂ nanostructures under realistic thermal boundary conditions. Model validation is performed through comparison with experimental data from the literature, showing excellent agreement. The DFT analysis reveals that doping MgH₂ nanomaterials with Ti, V, and Zr modifies their thermodynamic properties, including enthalpy of formation and desorption temperature. At the reactor scale, these modifications lead to enhanced hydrogenation kinetics and improved thermal management. Compared to pristine MgH₂, hydrogenation time is reduced by 21%, 40%, and 42% for Ti-, Zr-, and V-doped nanomaterials, respectively, while thermal energy consumption during hydrogenation decreases by ~17% for V doping. These results highlight the strong correlation between nanoscale modifications and macroscopic system performance. The proposed multi-scale model provides a powerful tool for guiding the design and optimization of advanced nanostructured hydrogen storage materials for sustainable energy applications. Full article

With the advancement of image processing techniques, few-shot learning (FSL) has gradually become a key approach to addressing the problem of data scarcity. However, existing FSL methods often rely on unimodal information under limited sample conditions, making it difficult to capture fine-grained differences between categories. To address this issue, we propose a multimodal few-shot learning method based on category name expansion and image feature enhancement. By integrating the expanded category text with image features, the proposed method enriches the semantic representation of categories and enhances the model’s sensitivity to detailed features. To further improve the quality of cross-modal information transfer, we introduce a cross-modal residual connection strategy that aligns features across layers through progressive fusion. This approach enables the fused representations to maximize mutual information while reducing redundancy, effectively alleviating the information bottleneck caused by uneven entropy distribution between modalities and enhancing the model’s generalization ability. Experimental results demonstrate that our method achieves superior performance on both natural image datasets (CIFAR-FS and FC100) and a medical image dataset. Full article