Search Results (1,201)

This work addresses the need for efficient and reusable vector embeddings (VEs) for large-scale RDF Knowledge Graphs, focusing on the widely used DrugBankdataset. The study aims to reduce the computational burden and environmental impact associated with repeatedly generating embeddings for downstream tasks such as link prediction, clustering, and recommendation. To this end, embeddings are generated from the DrugBank Knowledge Graph—comprising over 3.6 million triples, more than 1.5 million entities, and 95 relations—using 25 models implemented in the PyKEEN framework. The dataset is processed into training, validation, and test splits, and embeddings of fixed dimensionality are produced for both entities and relations. The resulting representations are released in multiple formats, including full JSON files, class-partitioned subsets, and an efficient Parquet-based structure, to support scalable querying. Experimental evaluation demonstrates substantial improvements in access time and memory usage when using structured formats, particularly Parquet. Additionally, the study quantifies the carbon footprint of embedding generation, showing that distributing precomputed embeddings can reduce energy consumption by up to 99.37% compared to on-demand recomputation. Overall, the work provides a comprehensive, reusable resource that facilitates research while promoting computational efficiency and environmental sustainability. Full article

Spaceborne Global Navigation Satellite System Reflectometry (GNSS-R) archives usually provide only one delay–Doppler map (DDM) for each recorded observation condition, limiting the representation of residual DDM variability. This study proposes a Position-Guided Conditional Normalizing Flow (PGCFlow) for observation-grounded probabilistic expansion of ocean bistatic radar cross section (BRCS) DDMs. PGCFlow uses four invertible affine coupling blocks to map a 17 × 11 DDM to an equal-dimensional Gaussian latent space. Wind–Auxiliary Condition Modulation incorporates a seven-dimensional condition vector into affine-parameter prediction, while Position-Guided Cross-Partition Aggregation (PGCA) uses deterministic grid descriptors to retain explicit cell locations and facilitate spatial-dependence modeling. Experiments used 5,819,042 quality-controlled CYGNSS observations from 2024. PGCFlow was compared with a conditional variational autoencoder and a generic conditional invertible neural network on 8000 held-out recorded conditions drawn from the same empirical observation domain, with 16 generated DDMs per condition. Although the cVAE achieved the highest balanced-aggregate structural similarity (SSIM) of 0.9396, PGCFlow obtained the lowest Fair Energy Score (FES) and Variogram Score (VS) of 0.2242 and 0.0641 and the closest relative local-neighborhood dispersion to unity at 1.0501. It also achieved the lowest frozen-estimator response RMSE and response MAE of 1.1900 and 0.9129 m/s, respectively. Ablation results indicated individual contributions from both proposed modules. Overall, PGCFlow achieved a favorable trade-off among the evaluated fidelity, dependence, dispersion, and response-consistency measures. Full article

Distributed acoustic sensing (DAS) converts fiber cables into dense strain-rate sensor arrays, capturing direct, reflected and guided seismic phases with ultra-fine spatial sampling. However, DAS interrogators suffer far stronger random noise than traditional geophones, limiting its microseismic imaging capacity. This work adopts curvelet-transform denoising to suppress noise. Curvelets partition the frequency–wavenumber plane into multiscale directional sectors; coherent wave energy concentrates in limited angular wedges, while stochastic noise disperses evenly across all transform coefficients, enabling noise-signal separation via wedge-wise thresholding. We test four default threshold schemes on synthetic downhole DAS microseismic data. Parameter tuning proves all methods deliver comparable performance, so we compare their out-of-box reliability for shale reservoir monitoring. Three noise-statistic-based strategies perform stably: median-absolute-deviation (MAD), quiet-window and empirical-cumulative-distribution-Function (ECDF percentile) thresholding. By contrast, the default knee-point algorithm from mainstream DAS toolboxes fails, as its preset threshold falls within noise components and barely removes interference. We propose MAD as a robust default for the tested downhole DAS microseismic setting for it estimates thresholds directly from noisy traces without blank reference windows and offers superior operational stability. Applied to field DAS records from a southwest China shale-gas horizontal monitor well, the MAD curvelet workflow greatly enhances microseismic arrivals with negligible spurious events. Benchmarks against standard 2D Daubechies-4 wavelet and adaptive Goldstein FK filtering verify curvelet denoising as a physically interpretable, efficient tool for DAS wavefields. Full article

In long-aperture forward-looking radar, nonlinear Doppler mismatch caused by target relative motion can lead to positioning deviation and image defocusing. To address this issue, an adaptive segmented Doppler compensation method based on phase error constraints is proposed. As synthetic aperture time increases, high-order terms in the slant range history broaden the Doppler spectrum and enhance spatially variant phase errors. Conventional global compensation cannot achieve stable focusing, and fixed-length segmentation fails to adapt to varying motion nonlinearity. Accordingly, the high-order nonlinear characteristics of the slant range are first analyzed, and an adaptive sub-aperture partitioning criterion constrained by second-order phase error is derived, ensuring each sub-aperture satisfies the local quasi-linear hypothesis. A cross-segment mapping relationship between different sub-apertures is then established, and the compensation process is formulated as a two-dimensional separable operator. To manage the high computational complexity of solving spatially variant mapping under long apertures, the Alternating Direction Method of Multipliers (ADMM) is introduced to iteratively optimize the operator, achieving phase alignment and coherent reconstruction among sub-apertures. Simulation and experimental results show that the proposed method effectively suppresses nonlinear defocusing under long-aperture conditions. Compared with conventional global methods, it achieves superior energy concentration and focusing resolution in extended target scenarios. Full article

A comprehensive Quasi-Classical Trajectory (QCT) investigation is reported for the hydrogen-abstraction reaction, OH (ν = 0, j = 2) + D₂ (ν = 0, j = 2) → HOD* + D, aimed at clarifying the mechanisms of energy and angular-momentum disposal in the excited product molecule, HOD*. Calculations were performed on the Wu–Schatz–Lendvay–Fang–Harding (WSLFH) potential energy surface for five collision energies, E_T = 0.28–0.64 eV. Vibrational and rotational Gaussian Binning (V-GB and R-GB) methods were applied to obtain state-resolved distributions of the relative translational energy, P(E_T′), and internal angular momentum, P(J′), of HOD*. The results show pronounced mode selectivity, with dominant excitation of the OD-stretch vibration at low collision energies, characteristic of an early-barrier abstraction mechanism. Increasing E_T promotes a gradual transfer of energy into translational and rotational motion, leading to broader and less anisotropic P(J′) distributions. The combined V-GB and R-GB analysis provides a consistent description of the coupling between translational, vibrational, and rotational degrees of freedom. These findings provide further insight into rovibrational energy partitioning in the OH + D₂ reaction and may contribute to the fundamental assessment of this system as a possible source of rovibrationally excited molecules for chemical-laser schemes. Full article

High penetration of distributed energy resources (DERs) creates opportunities for planned island formation and operation in distribution networks, which can help mitigate voltage fluctuations, reduce reverse power flow, and improve renewable energy utilization. This paper proposes a rapid planned island formation method based on a graph attention network (GAT) and a multi-layer perceptron (MLP). First, considering the technical operation requirements of distribution networks, the factors affecting island formation, including topological connectivity, power balance, and DERs, are analyzed. Based on graph theory, the topology connections and node information of the distribution network are represented by an adjacency matrix, a node feature matrix, and an edge physical feature matrix, respectively, to construct the distribution network graph model. Second, the GAT is employed to learn the correlations among distribution network nodes and obtain a node feature matrix incorporating relational information. The obtained representations are fed into the MLP to learn the mapping between node representations and branch connectivity states, and an island partitioning scheme is generated based on the output probabilities. Finally, considering the power balance constraints of the distribution network, the boundary nodes of the islands are adjusted to obtain the final feasible island partitioning scheme. Simulation results on a modified IEEE 33-bus distribution system show that the proposed method achieves an F1 of 92.37%. Furthermore, the model maintains an average F1 of 91.32% in generalization tests, demonstrating its effectiveness and robustness for rapid planned island formation in distribution networks with high penetration of DERs. Full article

This paper investigates joint computing and relaying for multiuser task offloading in a wireless-powered mobile edge computing (MEC) system comprising an energy node (EN), an edge server (ES), and multiple energy-harvesting users. One user is selected as the helper for the remaining task users. Each task user partitions its workload among local computing, cooperative computing at the helper, and remote execution at the ES. During a parallel cooperation stage, the helper computes one portion of the uploaded tasks locally while forwarding the remaining portion to the ES and also processes its own task through local computing or edge offloading. The weighted sum computation rate (WSCR) is maximized by jointly optimizing helper selection, task partitioning, time allocation, transmission-energy allocation, and CPU-resource allocation under frame-duration, energy-neutrality, communication, and computation constraints. For each candidate helper, transmission-energy variables are introduced to decouple transmission time and power, and the perspective structure of the achievable-rate functions is exploited to reformulate the continuous resource-allocation problem as an equivalent convex problem. By solving the convex problem for all the candidate helpers, the globally optimal helper selection and resource allocation are obtained. The numerical results show that the proposed joint computing-and-relaying scheme consistently outperforms computing-only, relaying-only, and dedicated-helper cooperation. The performance gain stems from adaptively balancing helper computing and ES processing according to the prevailing communication, computation, and energy bottlenecks. Full article

Floating offshore wind turbines (FOWTs) experience coupled aero–hydro–servo-elastic excitations that produce structurally distinct kinematic responses at different measurement heights. While field monitoring campaigns increasingly deploy multi-level inertial sensors, the quantitative spectral partitioning of response energy across measurement levels and its relationship to operational and environmental conditions remain poorly characterised for operational FOWTs. This study presents a systematic multi-level spectral decomposition of operational FOWT structural response using synchronised tower-base and nacelle strapdown inertial measurements acquired at 8 Hz over a six-day campaign (18–23 April 2023) at a semi-submersible FOWT in Chinese coastal waters. Six kinematic channels—three translational acceleration components and three translational velocity components—from each sensor are decomposed into four physically defined frequency bands: drift (0.005–0.05 Hz), wave (0.05–0.30 Hz), structural (0.30–0.50 Hz), and rotor (0.50–0.80 Hz). Three derived scalar metrics—band energy ratio (BER), Wave-to-Structural Dominance Ratio (WSDR), and Structural Amplification Factor (SAF)—are defined, with their complete computation specifications and parameter sensitivity analysis provided to ensure reproducibility. Across 36 ten-minute windows spanning diverse conditions (mean wind 4.2–12.1 m/s, $H_s$ 0.8–3.1 m), results reveal a pronounced and consistent spectral separation: the tower base is strongly wave-dominated (${\mathrm{BER}}_{\mathrm{wave}}$ = 75.9%, coefficient of variation CV = 15.0% across days), whereas the nacelle exhibits substantially elevated structural-band energy (${\mathrm{BER}}_{\mathrm{struct}}$ = 10.3%, 4.72-fold amplification relative to tower base, 95% CI [3.63, 5.81]) and rotor-band energy (11.8%, 3.77-fold amplification). The WSDR at the tower base (mean 200.9, 95% CI [101.4, 300.4]) exceeds that at the nacelle (mean 48.4, 95% CI [13.1, 83.7]) by a factor of 4.1×. One-way analysis of variance (ANOVA) reveals that nacelle structural-band BER is significantly modulated by SCADA operational regime ($F=5.20$, $p=0.024$, ${\eta }^2=0.16$) and by significant wave height ($p=0.031$), while tower-base wave-band BER is primarily driven by $H_s$ ($p=0.018$). Comparison with baseline features—root-mean-square acceleration, spectral peak frequency, and traditional broad-band energy ratio—demonstrates that the band-resolved BER provides finer discrimination between excitation mechanisms than aggregate metrics. Importantly, no structural damage events occurred during the monitoring period; therefore, the reported stability of these features is interpreted as a baseline characterisation under normal operational conditions, which could support future anomaly detection efforts but does not constitute validation of damage detection capability. A comprehensive limitations assessment is provided, covering single-turbine validation, frequency-band sensitivity, regime sample imbalance, and generalisability constraints. Full article

European Union countries can reach comparable electricity-generation mixes through transformations that differ in sequence, timing, and structural exposure. This study examines annual mixes for all 27 Member States in 2000–2024 as ten-part closed trajectories. The primary specification preserves reported zeros, applies a Hellinger transformation, compares trajectories by multivariate dynamic time warping with a four-year Sakoe–Chiba window, and clusters them using partitioning around medoids. Three conditional multi-country pathway types were identified among 26 countries: a diversifying fossil-to-renewables pathway, nuclear-renewables coexistence, and a concentrated island transition. Estonia was retained as a country-specific oil-shale profile outside the typology. Separation was moderate (mean silhouette = 0.2802), although bootstrap stability was high. Endpoint-only and full-trajectory partitions agreed only partly (Adjusted Rand Index = 0.4695), but a strict pairwise test found no extremely similar endpoints combined with extremely different paths. The representation of the mix reorganized the partition, whereas elastic time alignment changed only one assignment. Because imports, interconnection, flexibility, reliability, costs, and adaptive capacity lie outside the data, the typology does not measure overall energy-security risk. It shows instead that declining fossil shares redistribute rather than remove structural exposure, and it identifies where concentration, technology dependence, and substitution patterns warrant deeper assessment. Full article

Alkali metal immobilization mediated by kaolin during wheat straw gasification was investigated using thermodynamic equilibrium calculations. Kaolin promoted the formation of stable K- and Na-bearing aluminosilicates, including feldspar, leucite, and nepheline. The controlling mechanism shifted from mineral-phase reactions below 1100 °C to a molten slag structure at higher temperatures. Increasing kaolin addition reduced the slag structure parameter R from 0.43 to 0.09, indicating enhanced network polymerization. The calculated evolution of Al-containing network units was consistent with enhanced K⁺/Na⁺ charge compensation around tetrahedrally coordinated Al, suggesting a possible structural origin for the increased thermodynamic stability of alkali metals. Gibbs free energy calculations demonstrated the improved thermodynamic stability of K and Na in the slag phase. At a kaolin addition of 5 wt.%, the K release fraction decreased by at least 29.36 percentage points relative to wheat-straw ash even at high temperatures over 1400–1600 °C. The results clarified the thermodynamic relationships among ash composition, phase evolution, and alkali partitioning during kaolin-assisted gasification. Full article

The rapid proliferation of AI-enabled Industrial Internet of Things (IIoT) applications has significantly increased the energy consumption of shared edge computing infrastructures. Despite this sustainability challenge, contemporary resource pricing models in edge computing environments remain largely anchored in coarse-grained physical resource allocations rather than the actual energy consumed during workload execution. To support energy-aware resource management in industrial edge computing, this article formalizes an energy attribution model for AI workloads executed in shared edge nodes, where multi-tenant workloads concurrently share computing resources. The primary challenge in such environments stems from the inherent non-separability of localized power consumption among co-located workloads due to dynamic resource sharing and execution interference. To address this technical hurdle without introducing prohibitive monitoring or instrumentation overheads to the industrial environment, our model partitions aggregate system-level energy metrics into baseline platform elements and resource-specific functional components. It then formulates energy shares to individual workloads by solving consistent attribution functions based on observable resource allocation and utilization variables. By categorizing infrastructure hardware into utilization-driven, allocation-centric, and hybrid behavior profiles, the model precisely approximates workload-level energy responsibility within a practical error margin in shared IIoT edge computing environments. Full article

Surface energy partitioning regulates heat and water exchange between land and atmosphere and reflects alpine ecosystem responses to meteorological variation. Using radiation and meteorological data from November 2022 to October 2023, we compared adjacent alpine shrubland (Hulu 1) and alpine meadow (Hulu 2) ecosystems in the Qilian Mountains. Surface energy fluxes were estimated with a combined method based on surface energy balance, then evaluated with eddy covariance measurements. Path models examined direct and indirect environmental effects on turbulent fluxes. Standardized sensitivity coefficients based on evaporative fraction (EF) assessed seasonal responses of energy partitioning to environmental variation. Both ecosystems showed similar seasonal patterns, although flux magnitudes differed. Net radiation (R_n) followed a unimodal annual cycle and averaged 107.69 W m⁻² in the meadow and 89.36 W m⁻² in the shrubland. Sensible heat flux (H) peaked in May, with annual means of 60.22 and 51.68 W m m⁻². Latent heat flux (LE) peaked in July and averaged 49.12 and 38.58 W m m⁻². Soil heat flux (G) varied least, averaging −21.64 and −0.89 W m⁻². Path analysis identified R_n as the strongest control on turbulent fluxes. Its effect on H was weaker in the shrubland (0.92) than in the meadow (0.97), whereas its effect on LE was stronger in the shrubland (0.94) than in the meadow (0.71). Wind speed was positively related to H but negatively related to LE, with a stronger effect on H in the shrubland. Vapor pressure deficit (VPD) was negatively related to H but positively related to LE. Soil water content (SWC) had limited direct effects on turbulent fluxes at both sites. Sensitivity analysis showed higher overall EF sensitivity to environmental variation in the meadow during the growing season (0.510 vs. 0.228). Meadow EF was more sensitive to soil temperature (T_s) and SWC, whereas shrubland EF responded more strongly to VPD. Over the whole period, overall EF sensitivity was higher in the shrubland than in the meadow (0.439 vs. 0.353). These findings show that vegetation type and local environmental conditions jointly shape surface energy balance and energy partitioning in alpine ecosystems. Full article

Renewable-generation forecasting is a probabilistic time-series problem in which point accuracy alone is insufficient for operational decision-making. This study proposes a physics-informed Bayesian framework for multi-horizon forecasting of solar, wind, and total renewable generation. The task is formulated as a 15-dimensional target–horizon problem covering three generation families and five forecast horizons: H1, H3, H6, H12, and H24. A leakage-safe data construction protocol generates causal predictors from meteorological observations, generation history, calendar cycles, lagged and rolling statistics, ramp descriptors, and physics-informed transformations. PI-BHTF partitions the 422-dimensional predictor space into solar, wind, temporal-calendar, and cross-context components, encodes them through parallel nonlinear branches, and combines the representations using cross-energy gated fusion. Its neural core uses a heteroscedastic predictive head and Monte Carlo dropout to distinguish input-dependent aleatoric uncertainty from epistemic variability, whereas the final hybrid forecasts are calibrated using residual quantiles computed from a chronologically held-out validation segment. Across three prespecified random seeds, the full PI-BHTF achieved a mean MAE of 0.0797 ± 0.0007 on the internal chronological test and 0.0686 ± 0.0006 on locked external SCADA validation. Performance and calibration varied substantially across target–horizon tasks, including marked short-horizon external undercoverage for wind generation. Component-wise ablation supported semantic feature partitioning and the heteroscedastic head, whereas the physics-informed features, cross-energy gate, and physics-consistency loss did not independently reduce aggregate MAE. PI-BHTF should therefore be interpreted as a reproducibly competitive framework that balances multi-horizon accuracy, structured representation, uncertainty estimation, and external transferability rather than as a universally dominant model. Full article

Caffeine (1,3,7-trimethylxanthine) is a widely consumed psychoactive drug and neurostimulant, yet its molecular organization and permeation behavior in lipid membranes are not fully understood. We employ microseconds-long, united-atom molecular dynamics simulations to investigate caffeine interactions with a solvated DPPC (1,2-dipalmitoyl-sn-glycero-3-phosphocholine) bilayer at its fluid phase. Caffeine molecules initially aggregate in the aqueous phase to form ordered stackings, which successively permeate into the membrane. The stacked assemblies gradually dissolve, reaching a stable dispersed state where caffeine molecules preferentially reside near the headgroup–acyl chain interface and orient parallel to the lipid acyl chains, consistent with previous experimental and simulation studies. Simulations initiated with caffeine in the membrane hydrophobic phase converge to the same equilibrium state, indicating a preferred localization at the interface region. Present simulation findings are further supported by free-energy calculations that demonstrate caffeine’s high affinity at the headgroup–acyl chain interface. Caffeine partitioning transiently and slightly reduces bilayer thickness and increases the membrane surface area while enhancing acyl chain stiffness near the hydrophilic part of the membrane. These observed trends are reproducible over different system sizes and independent simulations. Overall, this study provides atomic-level insights into the caffeine permeation process, including its effect on the lipid bilayer structure. Full article

Energy efficiency is a primary design objective for battery-powered IoT devices. While offloading computation-intensive tasks to edge servers has been extensively studied to mitigate power drains, relatively little attention has been paid to the long-term financial cost of commercial edge services. This article proposes Volunteer Edge, a cost-effective real-time task offloading framework that exploits underutilized computing resources of privately managed nodes to execute offloaded workloads. Unlike conventional public edge servers, volunteer edge nodes provide inexpensive computing resources but are subject to unpredictable node churn. To address this challenge, we present a dual-class task model that partitions workloads into critical and normal tasks, and selectively applies task replication to volunteer edge nodes. The framework jointly optimizes task placement, processor frequency scaling, and replication decisions using a steady-state genetic algorithm to minimize task execution cost and IoT-device energy consumption while satisfying schedulability and reliability constraints. Extensive simulations demonstrate that Volunteer Edge significantly reduces offloading cost while maintaining IoT-device energy efficiency and protecting critical tasks against volunteer node failures. Specifically, the proposed framework reduces edge rental costs by 54.0% on average compared with conventional public-edge-based offloading while maintaining reliable execution of critical real-time tasks. Full article