Search Results (858)

Long-term time series forecasting (LTSF) is critical for modern power systems, energy management, and grid planning. Yet virtually all existing forecasting models employ stationary activation functions that apply identical nonlinear mappings regardless of temporal context—a fundamental mismatch with real-world load data, which exhibits strongly regime-dependent dynamics such as summer demand peaks, winter heating patterns, and overnight low-load periods. We address this gap by proposing TC-KAN (Time-Conditioned Kolmogorov–Arnold Network), the first forecasting architecture to augment KAN activation functions with position-aware coefficient parameterisation. The core innovation replaces the static polynomial coefficients in standard KAN activations with position-conditioned coefficients produced by a lightweight positional-embedding MLP, providing additional learnable capacity beyond standard KAN while adding negligible parameter overhead. TC-KAN further integrates a dual-pathway processing block—combining depthwise convolution for local temporal pattern extraction with the time-conditioned KAN layer for enhanced nonlinear transformation—within a channel-independent framework with Reversible Instance Normalisation. Experiments were conducted on four standard ETT benchmark datasets and the high-dimensional Weather dataset. TC-KAN achieves superior or competitive accuracy in most configurations while requiring merely 51K parameters—approximately 40% of DLinear and ∼100× fewer than iTransformer. On ETTh2, TC-KAN reduces the mean squared error by up to 61.4% over DLinear, and matches the current state-of-the-art iTransformer on ETTm2 at a fraction of the computational cost. This extreme parameter reduction circumvents the steep memory bottlenecks endemic to massive Transformer models, positioning TC-KAN as a highly practical architecture tailored precisely for resource-constrained edge deployments—such as on-device load forecasting inside smart grid sensors and industrial IoT controllers. Full article

Microfluidics enables spatially controlled nanostructure synthesis by coupling confined flows with surface reactions. In this work, we study how geometry-induced laminar microenvironments govern the in situ formation of Au and Ag nanostructures inside 3D-printed microfluidic reactors. Proof-of-concept fish-scale valves were fabricated by masked stereolithography in three architectures designed to define three recurring zones in the microreactor, inside the fish-scales (zone 1), between the fish-scales (zone 2), and along the rows of fish-scales (zone 3). A Cu thin film was deposited on the inner walls of the channel to serve as the sacrificial surface for galvanic replacement using AgNO₃ or HAuCl₄. Distinct 0D, 1D, and 2D nanostructures were simultaneously obtained in a zone-dependent manner across the valves, including nanoparticle and nanopore-rich regions, nanowires, nanoflakes and clustered 2D features. COMSOL simulations were used to solve the Navier–Stokes equation and extract specific-zone flow descriptors, including Reynolds number, velocity, and wall shear stress, and relate them to the nanostructure morphologies observed by SEM. The flow throughout the devices is strongly laminar, with local Reynolds numbers up to 0.04, exhibiting systematic spatial gradients imposed by the valve geometry. These results provide a design-guided route to tune nanostructure morphology through microchannel architecture under constant global operating conditions. Full article

A numerical study was conducted to quantify the entropy generation in a fluidic oscillator operating at Reynolds numbers of 30,000, 40,000, and 50,000. Both the local entropy production rate and total entropy were calculated under these operating conditions. Transient computational fluid dynamics (CFD) simulations were carried out using the $k-\omega$ shear stress transport (SST) turbulence model. The total entropy was compared with the pressure and driving-force coefficients to establish its relationship with force dynamics. The total entropy showed a periodic evolution synchronized with the jet switching process, while its amplitude increased with Reynolds number and showed a slight phase delay. The pressure and driving-force coefficients exhibited weak fluctuations at the end and beginning of each oscillation period, matching the secondary peaks in total entropy and indicating that these variations arise from residual dissipative effects linked to the jet reattachment stages. The local entropy production rate was concentrated near the feedback channels, Coanda surfaces, and the interaction zone where the jet from the inlet nozzle met the returning flow from the feedback channels. Regions of elevated entropy were detected at the outlet corners due to expansion and pressure drop. The high-velocity jet core exhibited minimal entropy, which increased toward the flanks as the flow decelerated. The results show that entropy generation follows the jet switching motion, reflecting the variations in viscous dissipation and flow dynamics inside the oscillator. Full article

Shock-aware aerodynamic shape optimization of transonic airfoils requires surrogate models that capture both integral aerodynamic trends and shock-relevant pressure distribution features. This study addresses drag-oriented optimization of the RAE2822 transonic airfoil under a lift-targeted condition with baseline relative thickness feasibility, rather than strict target pressure inverse design. Each airfoil is parameterized by a 16-dimensional CST vector and mapped to a two-channel vertical signed distance field representation of the upper- and lower-surface $C_p$ curves, from which shock descriptors, including the shock location indicator $x_s$ and the pressure jump magnitude $\Delta C_p$, are extracted in a deterministic, implementation-consistent manner. To quantify the reliability of surrogate-derived shock metrics, a held-out uncertainty analysis is performed on 500 samples. The surrogate achieves MAE/RMSE values of 0.00474/0.00602 for $C_L$ and $4.66\times {10}^{-4}/6.33\times {10}^{-4}$ for $C_D$, while the recovered shock-related quantities yield 0.00201/0.01598 for $x_s$ and 0.00200/0.00336 for $\Delta C_p$. Scatter plots and error histograms show tight one-to-one trends for most samples, with limited outliers mainly associated with locally ambiguous pressure gradient patterns. Overall, the surrogate is more reliable for capturing shock intensity trends than for prescribing an exact shock location; accordingly, $x_s$ is interpreted as a trend-level descriptor, whereas $\Delta C_p$ is treated as the more stable engineering indicator inside the optimization loop. The trained surrogate is embedded in a differential evolution optimizer with soft penalties on lift deviation and thickness feasibility violation, and selected designs are re-evaluated through closed-loop SU2 RANS simulations. CFD verification shows that the optimized design reduces drag from $C_D=0.01463$ to $C_D=0.01229$ (a 16.0% reduction) and reduces the shock jump from $\Delta C_p=0.239$ to $\Delta C_p=0.046$ (an 80.7% reduction). For the optimized design, the prediction-to-CFD differences are $\Delta C_L=+0.0042$ and $\Delta C_D=+0.00012$. These results support an engineering-oriented and auditable shock-aware closed-loop optimization workflow, with final design conclusions established by CFD verification rather than surrogate-predicted shock location alone. Full article

Line spectrum purification is a fundamental task in underwater detection and identification tasks. A dual-input architecture based on Dense U-net is introduced to extract clean line spectra from strong interference. The U-net model features a symmetric encoder–decoder structure that accepts two-dimensional data as both input and output. The DenseBlock, a core component of DenseNets, offers greater parameter efficiency compared to conventional convolutional layers. In this paper, standard convolutional layers inside the original U-net are replaced by DenseBlocks. This model possesses two input channels, thus allowing the time–frequency feature of the interference and that of the interference–target mixture to be fed simultaneously. With supervised learning, the model is capable of eliminating the strong interference components and background noise from the superimposed spectrum, thereby producing a purified target line spectrum. Compared to traditional interference suppression methods, this approach offers higher feature accuracy and greater signal-to-interference-and-noise ratio (SINR) gain. Moreover, the model is trainable using simulation datasets and then deployed to real-world measurements, demonstrating strong generalization capabilities—a valuable property given the limited availability of labeled samples in underwater detection tasks. Being data-driven, this method operates without requiring prior assumptions about the array configuration, and consequently exhibits greater resilience to array imperfections relative to conventional model-based interference suppression techniques. Simulation and experimental results demonstrate that the proposed method achieves an output SINR improvement of more than 8 dB under low SINR conditions and exhibits significantly better robustness to array position errors than conventional methods, verifying its excellent line spectrum purification capability. Full article

Driven by the large-scale deployment of renewable electricity, water electrolysis has emerged as a leading pathway for high-efficiency hydrogen production. Under practical operating conditions, gas evolution and gas–liquid two-phase flow inside electrolyzers substantially reshape electrode interfacial states and the in-cell mass transfer environment and have been reported to cause performance losses on the order of 10–30% under unfavorable conditions. This review summarizes the evolution of electrode-generated bubbles during nucleation, growth, detachment, and coalescence, and consolidates the fundamental features of two-phase hydrodynamics and phase-distribution patterns in electrolyzer channels. Progress and limitations of major two-phase modeling approaches are then assessed with respect to their capability to resolve the relevant interfacial and transport processes. The impacts of gas evolution and two-phase flow on electrochemical performance, stability, and durability are subsequently discussed. Finally, recent advances in two-phase-flow management—through flow-field organization and structural design, as well as the introduction of external physical fields—are reviewed, together with experimental and diagnostic methods used to quantify bubble behavior and phase distributions. This review aims to provide a coherent understanding of the governing behaviors, research tools, and performance implications of gas evolution and two-phase flow in water electrolysis, and to inform electrode/transport-layer design, flow-field management, and the development of predictive numerical models. Full article

To investigate the effects of distribution characteristics of microscopic voids (including the connectivity degree, pore-throat morphology, and size) on the permeability performance of open-graded friction course (OGFC) asphalt mixtures with steel slag as the anti-skid wearing course, two-dimensional computed tomography (CT) images of OGFC steel slag asphalt mixture specimens were first obtained via X-ray technology. The MATLAB R2022b-based image subtraction algorithm was then adopted to identify the interconnected voids inside the specimens to quantitatively characterize the morphological differences in interconnected voids in OGFC steel slag asphalt mixtures with different gradations. Furthermore, Finite Element simulation by ANSYS 2021 R1 was conducted to explore the influences of the diversion angle of interconnected voids on the water flow characteristics of OGFC steel slag asphalt mixtures, involving the variation laws of water flow velocity, water pressure and flow path in the diversion structure, thereby analyzing the resultant effects on the permeability performance of the mixtures. The results show that the combination of X-ray CT scanning and image processing technology enables more convenient, accurate and intuitive characterization of the internal void distribution characteristics of the mixtures. It was found that the pore-throat properties, including size, length, quantity and equivalent diameter, are the dominant factors restricting the permeability capacity of OGFC steel slag asphalt mixtures. As the diversion angle increases from 20° to 60°, the pressure gradient increases by up to 103.92%. After passing through the diversion section, the flow velocity increases by approximately four times. The streamline density at the channel axis is 4.2–4.5 times that near the channel wall. This study realizes the rapid extraction of void characteristics and the identification of key influencing factors on the permeability performance of OGFC steel slag asphalt mixtures, an achievement that cannot be attained by the previous macroscopic research on the permeability performance of such mixtures. Full article

The Advanced Meteorological Imager (AMI) on GEO-KOMPSAT-2A (GK-2A) lacks a 2.1 μm shortwave infrared channel, precluding the dark target surface reflectance estimation that other geostationary aerosol retrievals rely on. We propose an improved land aerosol optical depth (AOD) retrieval in which background surface reflectance (BSR) is derived entirely from pixel-level bidirectional reflectance distribution function (BRDF) inversion using the scaled Ross-Thick Li-Sparse (sRTLS) kernel model fitted to geostationary time-series observations. Unlike existing approaches, the algorithm inverts the BRDF independently at each retrieval channel without relying on spectral reflectance relationships or external surface reflectance products; it assumes a low-background AOD during an initial accumulation period and then iteratively refines both BRDF coefficients and AOD. Two aerosol models—generic and dust—are supported, with a geographic dust-zone mask activating two-model selection during spring. Validation against 74 Aerosol Robotic Network sites over 2023 yields R = 0.86, RMSE = 0.15, and bias = −0.02, compared with R = 0.59, RMSE = 0.25, and bias = −0.04 for the National Meteorological Satellite Center (NMSC) GK-2A AOD product. The largest improvements appear at AOD ≤ 0.1 (bias: +0.03 versus +0.11) and AOD > 0.8 (bias: −0.12 versus −0.85). The full March–May (MAM) evaluation yields bias = −0.06 across all 74 sites. As a separate parallel retrieval restricted to matchups inside the geographic dust-zone mask, the proposed algorithm (dust model included) gives bias = −0.03, which worsens to −0.11 when only the generic model is applied—nearly a fourfold increase. A comparison against Himawari-9/Advanced Himawari Imager (AHI)—a co-located geostationary sensor carrying a 2.3 μm shortwave infrared (SWIR) channel—shows that the proposed algorithm (R = 0.897) outperforms Himawari-9/AHI (R = 0.855) across all metrics, demonstrating competitive accuracy without relying on a SWIR channel. Full article

The development of renewable energy and the increasing demand for electricity underscore the importance of pumped storage for grid stability. Under low-flow pump operating conditions, pump-turbines frequently exhibit hump characteristics, causing severe hydraulic instability and strong pressure pulsations. This study investigates the formation of a hump using full-channel numerical simulations based on the Scale-Adaptive Simulation turbulence model. The numerical flow–head characteristics were validated against the available experimental H–Q data, while the pressure pulsation results were used for qualitative mechanism analysis. The results reveal three major mechanisms: pre-swirl and spiral backflow in the draft tube, non-uniform runner inflow, and vortex flow-induced separation in the wicket gates. An analysis of entropy production reveals that vortex dissipation is responsible for as much as 71% of hydraulic losses in the hump region. In order to mitigate these effects, four stabilizing fins were installed inside the draft tube. The simulations indicate that the fins possess the capability to inhibit swirl and backflow, confine the vortices within the fin–runner interface, improve inflow uniformity and reduce overall hydraulic losses. As a result, the structural modification significantly attenuates the pressure pulsation amplitudes at key monitoring points and visibly shortens the recovery periods. The region of the hump and positive slope of the performance curve are considerably reduced while the head near the region of the hump is increased. Although the intrinsic hump characteristic is still present, the fin-based flow-control strategy can effectively improve the performance and stability of the pump-turbine, which can guide the design and optimization of high-efficiency pumped-storage plants. Full article

Several studies have investigated counterflow and concurrent flow in channels separated by a membrane to simulate mass transfer through membranes; however, few of them have used computational fluid dynamics (CFD). The current study aimed to numerically simulate and physically describe the distribution of pressure and velocity in counter-current flow by solving Navier-Stokes (N-S) equations in the channel and membrane pores (vertical channels). This is in contrast to most previous studies, in which the channel flow was simulated using N-S equations while ultra-filtration membrane flow was simulated using Darcy’s law. Consequently, the current study was executed using a CFD simulation to achieve several significant features: avoiding the execution of experimental tests, reducing the effort of model design and the expense and time consumption of fabrication, and facilitating the easy observation of variations in the pressure and the horizontal and vertical velocity for each point in the model. Two-dimensional CFD methods directly simulated the flow in channels and membrane pores to solve the N-S equations for each point in the whole domain, for which the velocity (horizontal and vertical) and pressure were calculated. In the current study, it was found that the pressure decreased from the inlet to the outlet of the channel, the horizontal velocity decreased from the inlet to the middle of the channel length and then increased to the outlet of the channel, and the vertical velocity decreased from the inlet to the middle of the channel length (L/2) with an upward direction (positive) and from L/2 to the outlet of the channel with a downward direction (negative). The analytical solution (1D model) was used to validate a numerical simulation (CFD) for the current study, but there were slight differences in the results between them. The results were perfectly explored and displayed the flow distribution patterns inside the channels and the membrane pores (vertical channels). The current study model represents the hemodialysis process. Full article

This study compares two policy instruments for decarbonizing China’s seafood exports to the EU and UK over 10 years using a recursive dynamic computable general equilibrium model. One instrument applies tariff-like carbon surcharges on embedded emissions at the border. The other recognises certified low-carbon production through tiered rate reductions or exemptions. The model constructs product-level carbon cost wedges for processing electricity, aluminium packaging, and cold-chain operations, then transmits them to border prices through pass-through and to import volumes through Armington demand. These mechanisms operate inside a dynamic setting with capital accumulation, learning, and technology adoption. We evaluate processed tuna, shrimp, whitefish, and fresh tilapia to reflect differences in energy use, packaging intensity, and cold-chain reliance. Results show that certification, especially when paired with targeted domestic green finance or tax offsets, speeds adoption of cleaner power and refrigerants and preserves market share better than uniform surcharges. Effects differ between coastal and inland production hubs, supporting location-specific policy bundles. Sensitivity analysis varies carbon prices, adoption speeds, and certification coverage within stated parameter ranges. We report trade, export revenue, emissions, investment, and welfare outcomes and identify product and channel drivers of exposure. Full article

Thermal stratification in deep reservoirs can cause ecologically problematic cold-water releases, and many existing selective-withdrawal phenomena rely on a limited set of fixed intake levels, which constrains their ability to follow seasonal shifts in the thermocline. Stepless stratified intakes with continuously adjustable flap gates offer quasi-continuous control of withdrawal depth, but their multi-gate, multi-brace layouts generate complex internal hydraulics whose energy-loss mechanisms are not well captured by conventional head-loss and resistance-coefficient metrics. In this study, physical-model measurements are combined with a validated three-dimensional numerical model, and entropy-production theory is used as a diagnostic to resolve where and by which mechanisms mechanical energy is irreversibly degraded inside a single-unit stepless stratified intake. The analysis shows that turbulent entropy production accounts for more than 98% of total dissipation, concentrated mainly in the flow channel and gate shaft, while the reservoir and outlet pipe contribute only weakly. Local entropy-production-rate fields indicate that dominant irreversibilities are associated with flow turning at the active gate leaves and with separation and wake development around horizontal and vertical braces, which generate low-velocity bands across gate levels and a low-velocity corridor in the shaft. Five geometric modification schemes targeting gate-entrance shaping and brace layout are evaluated; a combined brace-alignment and edge-rounding configuration most effectively weakens dissipation hotspots, improves discharge sharing among gate levels and reduces total entropy production. These findings show that entropy-based diagnostics can complement traditional hydraulic indicators and provide effective guidance for the design and refinement of stepless stratified intake structures. Full article

This paper presents an end-to-end simulation methodology for evaluating package-induced signal integrity (SI) degradation in a peripheral component interconnect express (PCIe) 5.0 channel. By integrating package, printed circuit board (PCB), and add-in card (AIC) structures into a unified simulation flow, the proposed approach enables accurate assessment of system-level eye diagram degradation. Various package-level degradation factors, such as impedance mismatch, meander routing, and via stubs, are assumed and designed to analyze their individual and combined effects on insertion loss, intra-pair skew, and eye diagrams. Results show that even localized discontinuities inside the package propagate and compound through the end-to-end channel, causing a significant reduction in the eye diagram at the system level. These findings demonstrate that package-induced impairments cannot be evaluated solely at the package level but must instead be analyzed within a complete end-to-end channel environment. The proposed methodology provides a practical framework for predicting system-level SI degradation caused by package design choices, offering valuable insights for next-generation high-speed package and channel co-design. Full article

This paper investigates the different relaxation channels of a single symmetric top NH₃ and a spherical top CH₄ molecule trapped at low temperature in a clathrate hydrate nano-cage in the infrared absorption domain of their vibrational degrees of freedom. The approach utilizes the Born–Oppenheimer approximation and the extended site inclusion model applied to CO₂ in a previous work, which was based on pairwise atom–atom effective interaction potentials. The calculations show that trapping the methane or ammonia molecule is energetically more favorable in a type sI clathrate structure than in an sII one, and entropic considerations show that methane can be released much more easily than ammonia from clathrate hydrate nano-cages. In the small (s) and large (l) nano-cages with the sI structure, the CH₄ molecule exhibits a more or less perturbed rotational motion, while the NH₃ molecule shows a strongly hindered orientational motion that tends to a three-dimension librational motion (oscillation motion) around its orientational equilibrium configuration. The calculated orientational energy level schemes are quite different from those of the molecular free rotation. In the static field inside the cage, degenerate ${\nu }_3$ and ${\nu }_4$ vibrational modes of methane and ammonia molecules are shifted and split. Moreover, for ammonia molecules, the ${\nu }_1$ and ${\nu }_2$ modes are shifted, and the inversion motion is no longer allowed. The non-radiative and radiative relaxation channels of CH₄, NH₃ and CO₂ in clathrate nano-cages are discussed with reference to the matrix isolation spectroscopic results. Upon laser excitation, then, from the energy levels calculated for the different degrees of freedom, NH₃ and CO₂ are expected to fluoresce, while for CH₄, non-radiative relaxation should lead to evaporation at the surface of clathrates. Experimental setups are suggested to localize and study these species underneath ice surfaces on distant planets or planetesimals from mobile detectors such as drones or CubeSats equipped with appropriate laser sources and telescopes with 2D imaging detectors. Full article

Aquaporin 1 (AQP1) is a transmembrane protein that acts as a highly selective channel for the rapid passage of water across cell membranes, driven by osmotic gradients. The narrowest part of the water channel pore—the selectivity filter (SF)—plays a key role in ensuring selective and efficient water transport. In this study, density functional theory (DFT) at the M062X/6-311+G(d,p) level was used to identify the preferred position of the water molecule(s) inside the SF and to elucidate the forces that lead to its displacement during permeation. A systematic scan along the pore axis identified a well-defined energy minimum where a single water molecule was optimally stabilized by hydrogen bonds with SF residues. A second water molecule was introduced to study how the incoming water affects the translocation of the first water molecule. The resulting energy and force profiles reveal that the approaching water molecule gradually pushes the bound water forward, ultimately occupying its favorable binding site. These results provide an atomistic description of the positioning and displacement of water molecules in SF and offer a quantitative view of the fundamental interactions that govern water transport in AQPs. Full article