Search Results (5,764)

News recommendation is a critical technology that helps users efficiently find content of interest from large candidate pools. Its core objective is to accurately model user reading interests. However, current news recommendation systems typically suffer from two key limitations: (i) they fail to suppress noise from a frequency-domain perspective, and (ii) they lack effective calibration for popularity bias within the embedding space. In this work, we propose a novel frequency-domain denoising and popularity-bias correction recommendation network (FDDP-RN) to address both challenges simultaneously. Our approach introduces spectral analysis into the news encoder. Specifically, we design a filtering mechanism that combines truncation and scaling to enhance high-frequency semantic components, improve text feature representation accuracy, and suppress redundant low-frequency components. In addition, we introduce a norm-scaling factor that dynamically calibrates the embedding distribution of cold-start news items, placing them on an equal footing with popular news items. This effectively improves the exposure of long-tail content without requiring extra user interactions. We conduct extensive experiments on three public datasets, namely, MIND-small, MIND-large, and Adressa. The quantitative results demonstrate that FDDP-RN achieves state-of-the-art performance. Notably, on the Adressa dataset, our model achieves an AUC of 75.36% and an nDCG@10 of 50.11%, outperforming the strongest baseline. Furthermore, cold-start fairness diagnostics on the MIND-small dataset reveal that our mechanism increases the top-10 long-tail exposure rate from 15.3% to 18.1% and reduces the exposure Gini coefficient from 0.991 to 0.987, indicating a better balance among recommendation accuracy, diversity, and fairness. Full article

Heat pumps are core equipment for efficient low-grade thermal energy utilization and low-carbon transformation of the energy structure, offering significant energy-saving potential in building heating and industrial waste heat recovery. This paper reviews the research progress and technical challenges of compression, absorption, and adsorption heat pumps as well as nanofluid-enhanced heat transfer technology and elastocaloric heat pump systems. Air source heat pumps can delay frosting through variable frequency, heat storage, and waste heat recovery. However, accurate prediction models for performance degradation under extreme cold conditions are lacking. Although ground source and water source heat pumps exhibit significant energy efficiency advantages, ground source systems may suffer from performance degradation due to underground thermal imbalance. The application of water source systems is strictly constrained by water resource conditions. Driven by low-grade waste heat, absorption heat pumps employing traditional working pairs suffer from crystallization, corrosion, or high rectification energy consumption. The COP of a single-effect cycle under 80~100 °C waste heat is only 1.2~1.9, while hybrid cycles can reach approximately 3.2 at 120~150 °C. Although adsorption heat pumps achieve significantly improved performance under continuous heat recovery cycles, the full-scale power density of novel adsorbents such as metal–organic frameworks is inferior to the power density of traditional silica gel. Moreover, under off-design conditions, the performance drops by 23~48% compared to theoretical values. Nanofluids can enhance heat transfer, but the long-term effects of particle agglomeration at high temperatures on pump power consumption and system compatibility remain to be systematically evaluated. Elastocaloric heat pump systems can achieve refrigerant-free cooling, but current prototypes still cannot compete with traditional vapor compression systems in long-cycle fatigue reliability and power density. Current heat pump technologies generally face challenges such as insufficient adaptability to extreme conditions, bottlenecks in working fluids and materials, and a lack of long-term validation. Future research must construct a multi-source coupling optimization system, address common problems in working fluids and materials, promote long-term validation and kilowatt-level prototype demonstrations, and drive the large-scale deployment and engineering application of heat pump technology toward high efficiency, intelligence, and high reliability. Full article

Electrical power systems are exposed to interacting electrical, thermal, environmental, and resource-related faults such as leakage current, voltage and frequency deviations, overcurrent, harmonic distortion, phase-sequence error, humidity, fire, wind-speed variability, water insufficiency, and solar-resource loss. This study introduces a software-based database-generation and smart-learning framework that converts 22 candidate risk factors into six normalized severity levels and then maps the simultaneous system state to low-, medium-, and high-level protection decisions. The main novelty is that the software not only evaluates existing measurements; it also produces a literature- and standards-informed synthetic database when long-term real field measurements are not yet available. The database is generated by defining variable limits, sampling realistic operating states, computing severity labels, and storing input–output pairs that can later train or validate predictive maintenance models. The proposed framework therefore, links protection logic, database construction, and reusable training data in a single workflow. The results show how simulated annual operating scenarios can be transformed into structured risk records, warning classes, and shutdown decisions, supporting early fault detection, maintenance planning, and resilience improvement in renewable-integrated electrical networks. Full article

A double-ended traveling-wave accurate fault location method based on variational mode decomposition (VMD), generalized S-transform (GST), and a symmetric difference energy operator (SDEO) is proposed for single-phase-to-ground faults in small-current grounding distribution networks. The method is designed for fault conditions in which the fault current amplitude is low, the transient duration is short, and the initial traveling-wave wavefront is easily affected by high-frequency oscillation, reflection, refraction, and noise. The three-phase voltage traveling waves measured at both ends of the fault section are first transformed using Clarke modal transformation, and the voltage line-mode component is selected as the input signal. VMD is then used to decompose the nonstationary traveling-wave signal into several finite-bandwidth intrinsic mode functions. The high-frequency mode containing the initial wavefront mutation is processed using the generalized S-transform to enhance local time–frequency features. Finally, the SDEO instantaneous energy spectrum is used to calibrate the initial wavefront arrival time. For distance calculation, an improved double-ended traveling-wave location formula based on the horizontal section length and the absolute propagation time ratio at both line ends is constructed. This formulation reduces the dependence on a fixed empirical wave velocity and weakens the influence of line length deviation caused by practical line geometry. The method is verified using a deterministic PSCAD/EMTDC v5.0.2 and MATLAB R2021b co-simulation model with a fixed distribution network topology and arc-suppression-coil grounding. The tested cases cover selected fault distances, transition resistances, and fault inception angles. The simulation results show that the absolute location error remains within 100 m under all tested deterministic cases. The representative location error is 15 m at the 2.5 km fault point, 45 m under a 500 Ω transition resistance at the 1.5 km fault point, and 11 m under a 90° fault inception angle at the 6.15 km fault point. Compared with EMD-TEO, VMD-TEO, and VMD-GST, the proposed VMD-GST-SDEO method provides more stable wavefront calibration and lower location error in the studied cases. The results indicate the feasibility of the proposed approach within the stated simulation scope; further validation under stochastic noise, synchronization error perturbation, different sampling frequencies, and measured field waveforms is still required for engineering deployment. The fundamental novelty of the study lies in the task-oriented integration of VMD, GST, and the SDEO as a complete wavefront calibration chain and in coupling this chain with a propagation-time-ratio-based double-ended location formula for small-current grounding distribution networks, rather than in treating VMD, GST, or the SDEO as new standalone signal-processing algorithms. Full article

Urban areas are increasingly exposed to heat stress under climate change due to rising temperatures, ageing populations, and increasing population density in cities. This study assesses current and future heat-stress risk in Vilnius, Lithuania, using the Humidex index and the CLIMADA risk modelling framework. Historical heat-stress conditions are analysed using observational meteorological data, while future conditions are evaluated using five CMIP6 climate models under SSP2-4.5 and SSP5-8.5 climate scenarios for the mid and late 21st century. The risk assessment combines high-resolution Humidex-based hazard data, gridded population exposure, and age-specific vulnerability functions. In addition to temporal Humidex trend analysis, CLIMADA is used to estimate spatially explicit Expected Annual Impact and population-normalised impact indicators. The frequency of heat-induced stress has increased significantly—the number of days when the Humidex index was ≥30 has risen by 5.6 days per decade. It is projected that by the end of the century, the proportion of such warm-season days (April–October) will increase from 13.1% to 25.8% under the SSP2-4.5 scenario and to 41.6% under the SSP5-8.5 scenario. Spatial impact maps identify areas with elevated total and population-normalised heat-stress burden, while age-specific results show higher impacts among residents aged ≥ 65 years. Cumulative heat-stress burden, expressed as Expected Annual Impact, could increase by approximately 1.8–11.4 times. The framework provides a spatially consistent and age-sensitive approach for assessing urban heat-stress risk and supporting local climate adaptation planning. Full article

The complex and nonlinear load characteristics of Hall thrusters remain a key challenge in the design of propulsion power-supply output stages. In existing power-supply simulations for electric propulsion systems, the Hall thruster is often simplified as a fixed impedance or a prescribed current source, which makes it difficult to capture the time-synchronized interaction during simulation between the power-supply output stage and the thruster discharge process. To address this issue, this study encapsulates a one-dimensional discharge model as an externally callable thruster slave and proposes a HallThruster.jl–Simulink–Saber co-simulation method. The proposed method enables synchronized bidirectional exchange between the power-port voltage $V_{cmd}$ and the thruster discharge current $I_{out}$. The results show that the discharge current under the co-simulation condition exhibits a sustained low-frequency response at approximately 15.0 kHz. Compared with a fixed-voltage standalone simulation, the co-simulation preserves the same principal oscillation band and overall internal-field structures, while small but observable differences remain in instantaneous phase, local waveform shape, harmonic amplitudes, and high-gradient regions of the internal fields. The proposed method provides a computational framework for investigating dynamically coupled port behavior between a Hall thruster and a representative power-supply output stage. Full article

To address the sampling misalignment and detail loss caused by fixed-grid downsampling for small-scale defects, as well as the insufficient differentiated modeling and interaction of defect details and structural context in UAV-acquired power-line images, an enhanced lightweight YOLOv8n-based framework for power-line defect detection is developed. First, a content-adaptive downsampling (CAD) module is developed to predict input-dependent sampling offsets and normalized aggregation weights and to perform differentiable resampling. Combined with local-global interactive depthwise separable convolution, CAD improves the preservation of small-object details while maintaining relatively low computational complexity. Second, a dynamic subspace gated exchange (DSGE) module is proposed to adaptively partition features into a high-frequency detail subspace and a low-frequency structural subspace according to the input content. Heterogeneous branches and bidirectional gated exchange are then employed to jointly model fine-grained details and structural context. In addition, the lightweight mixed local channel attention (MLCA) mechanism is incorporated in the detection head as an auxiliary feature-enhancement component. Experimental results show that the proposed model achieves mAP@0.50 and mAP@0.50:0.95 values of 92.3% and 62.9%, respectively, outperforming the compared models under the current evaluation protocol. With 1.90 M parameters and 5.6 G FLOPs, the model reaches an inference speed of 134.7 FPS on the desktop GPU platform, demonstrating that content-adaptive sampling and dynamic detail–structure interaction can improve small-defect detection and complex-background suppression while maintaining relatively low model complexity. Full article

The energy transition entails numerous challenges related to the generation and distribution of various forms of energy and heat. In the case of electricity, these include, among others, challenges associated with transmission; however, these are usually less severe than those related to heat transmission, particularly due to limited access to district heating transmission networks. At the same time, heat storage is considerably less expensive than electricity storage, which means that heat management and balancing are also less costly than balancing electricity generation and demand. These two factors provided the rationale for selecting mobile heat transport from locations where surplus heat is generated to potential end-users as the subject of analysis. In this context, analytical methods were applied to determine the relationship between the economic efficiency of such a solution and selected technical and economic parameters, including CAPEX, OPEX, heat transport costs, and heat prices. The analysis demonstrated that, under current price conditions as of 2025, the economic efficiency of this solution, expressed using the net present value measure, may exceed zero only under specific conditions. These include relatively short transport distances, a high frequency of transport cycles, and favourable heat prices. Nevertheless, in the future, due to potential increases in heat prices or the possibility of subsidising this type of heat transport, particularly with respect to CAPEX, the solution may become economically attractive. Full article

Gout is a systemic metabolic inflammatory disease driven by hyperuricemia and monosodium urate crystal deposition, and is frequently accompanied by chronic kidney disease and uric acid nephrolithiasis. Current evidence indicates that acidic urine, hypocitraturia, and insufficient ammonium excretion are common in gout patients, and may be associated with insulin resistance, impaired renal ammoniagenesis, and abnormalities in tubular acid-base regulation. Persistent urinary acidification may contribute to the development and progression of uric acid stone formation and gout-related renal injury by reducing uric acid solubility, promoting crystal formation and intratubular deposition, and decreasing renal uric acid clearance. Urinary alkalization therapy can increase urine pH, thereby enhancing uric acid solubility and excretion, and has shown potential in some studies to improve serum urate levels, proteinuria, renal function parameters, and gout flare frequency, especially when using citrate-based alkali. However, the currently available evidence is mainly derived from observational studies and small prospective investigations, and recommendations across international guidelines remain inconsistent. High-quality evidence is still lacking regarding the optimal target population, urine pH range, choice of alkalizing agents, monitoring strategies, and long-term efficacy. Well-designed prospective studies are therefore needed to clarify the clinical role of urinary alkalization in renal protection among patients with gout. Full article

Standard actuarial methods often fail to accurately price the risk of catastrophic events due to the characteristic low frequency and high severity of such occurrences. To address these limitations, two actuarial solutions that integrate current knowledge of extreme value theory (EVT) with the theory of non-life insurance are proposed. To address the extreme risk, the Gumbel distribution (solution E1) and the chi-squared distribution (solution E2) were applied. The applicability of the proposed insurance models against the risk of windthrow damage was demonstrated on Norway spruce (Picea abies L. Karst.) stands at a national scale within the Slovak Republic. The results reveal a crucial economic relationship between the insurance layers: the higher the reliability and cost of coverage in the primary model, the cheaper the subsequent secondary premium for risk of extreme damage becomes. For both assumed return periods of 20-year and 100-year extreme events, the solution E2 generates lower and more acceptable total insurance premiums. Subsequently, both models address effectively the foundational insurance concept of risk minimization through spatial diversification. Full article

The conventional carrier phase-shifted pulse-width modulation strategy may cause asymmetric output phase voltages in a modular multilevel converter (MMC)-based static var generator, leading to high-frequency variations in the system common-mode voltage(SCMV). The resultant common-mode currents exacerbate insulation degradation and reduce equipment lifespan. To address this issue, this paper proposes a common-mode voltage (CMV) suppression method based on dynamic correction of the number of inserted submodules. By imposing the zero-SCMV condition as an insertion-number constraint, the proposed method dynamically corrects the numbers of inserted submodules in the three-phase arms, thereby suppressing the high-frequency SCMV fluctuations caused by the modulation process. Furthermore, integrating circulating current suppression techniques reduces the double-line-frequency alternating-current components components in inter-phase circulating currents, thereby enhancing the overall operational performance of MMC. Experimental validation on an RT-LAB-based rapid prototyping platform confirms the proposed method effectively mitigates high-frequency CMV variations. The results also show that this SCMV reduction is accompanied by increased output voltage total harmonic distortion (THD), which suggests that the proposed method is a compromise suppression strategy rather than a fully optimized modulation approach. Full article

Many kinases are known to homo- or hetero-merize and phosphorylate themselves or each other. Phosphorylation by itself (autophosphorylation) is mechanistically perceived as either cis- or trans-interactions. Identification of autophosphorylation sites in kinases provides an opportunity to assess the activity of the kinome in phosphoproteome datasets as an efficient indicator of kinase activity. In order to predict kinase activity-associated autophosphorylation sites, we first encompassed the currently reported homodimeric interactions between kinases from various sources. Subsequently, we accounted for the autophosphorylation sites in kinases that are predicted by phosphomotif screening approaches, in vitro kinase assays, and those perceived based on phosphomotif mutant analysis. Together, we identified 4184 autophosphorylation sites in 386 kinases, and among them, 315 kinases were reported with homodimeric protein–protein interactions. Furthermore, we analysed 361 kinases that harbor multiple autophosphorylation sites and 25 kinases with a single autophosphosite based on contemporary kinase–substrate interactions. Among the 386 kinases, autophosphosites for 17 kinases were derived through high-throughput or low-throughput approaches, and for 80 kinases, they were predicted based on their phosphomotif analysis. The remaining 289 were based on predictions by multiple tools or were reported in more than one dataset. There were many instances in which, although one auto-phosphorylation site was reported, multiple other sites were also predicted. In this regard, we analysed the instances of phosphosites in kinases that are already associated with the induction or inhibition of kinase activity and identified 223 kinases with one or more sites with functional association. To evaluate their potential as auto-phosphorylation sites, we systematically assembled 1572 profiles and 978 differential cellular phosphoproteome datasets. Subsequently, we accounted for the frequency of co-detection or co-differential regulation of these autophosphosites with multiple kinase activity-associated sites within each kinase and further accounted for their known/predicted substrate phosphorylation sites. Together, we provide a reference compendium of autophosphorylation sites secluded in the human kinome and their co-regulation with substrates based on global phosphoproteome datasets, AutoPhosDb. As indicators of kinase activity, these phosphosites in kinases would serve as a surrogate to enrich the active kinases and their potential substrate-specificity in phosphoproteome datasets. Full article

Magnetic stimulation modulates hippocampal synaptic plasticity in a parameter-dependent manner with effects shaped by stimulation frequency, intensity, waveform, and exposure conditions. Low-intensity magnetic fields generate induced electric fields that can influence neuronal membrane excitability and neural network activity. Hippocampal long-term potentiation (LTP) and long-term depression (LTD) are widely used as important readouts for evaluating the neurobiological effects of magnetic stimulation. Previous studies have shown that extremely low-frequency magnetic fields (ELF-MFs) may enhance, inhibit, or have no effect on LTP. These divergent effects appear to depend on stimulation conditions, developmental stage, and the baseline state of the neural network. High-frequency repetitive transcranial magnetic stimulation (HF-rTMS) has been reported to promote the recovery of LTP-like plasticity, improve synaptic structure, and regulate the expression of neurotrophic factors in some aging or pathological models. However, its effects are still influenced by both stimulation parameters and biological states. The underlying mechanisms may involve multiple levels of regulation, including Ca²⁺ dynamics, neurotrophic signals, glutamate receptor dynamics, mitochondrial function, and network oscillations. However, these mechanisms have been inferred mainly from various experimental models, and their interactions, temporal sequence, and causal relationships still need further clarification. Micro-magnetic stimulation (μMS) offers a potential technical approach for improving the spatial selectivity of local regulation in the hippocampus. Arrayed and wireless μMS platforms have further expanded their potential applications. However, available evidence has been obtained primarily from in vitro experiments and animal models. This review summarizes the effects of magnetic stimulation with different parameter configurations on hippocampal synaptic plasticity, integrates the biological mechanisms underlying cross-frequency magnetic neuromodulation, and discusses current challenges in the field of magnetic neuromodulation and future directions toward translational development. Full article

Roadside acoustic sensing may provide a low-cost approach to vehicle load-state monitoring. This study evaluates a normalized frequency-domain descriptor for controlled same-vehicle comparison. Vehicle-pass signals underwent resampling, pre-emphasis, framing, and windowing; event-level power spectral density from 500 to 5000 Hz was normalized and partitioned into 500–1000, 1000–3000, and 3000–5000 Hz bands. Field measurements comprised 60 passes of one 4 × 2 box truck at 20, 30, and 40 km/h with payloads of approximately 1000 and 4000 kg. Heavy load consistently increased the low-frequency proportion and decreased the high-frequency proportion. Two-way analysis of the additive log-ratio $ln(P_L/P_H)$ identified significant speed, load, and speed × load effects, with heavy–light separation decreasing as speed increased. A CarSim template was reparameterized at total masses of 9540 and 12,540 kg and a wheelbase of 7.15 m. Wheel-force impulses were convolved with a mass-dependent modal response to generate uncalibrated relative acoustic signals. None of the 48 sensitivity evaluations reproduced the full field direction. Thus, the descriptor supports same-vehicle comparison under controlled conditions, whereas the current dynamics-to-acoustic model is a reproducible discrepancy analysis rather than a calibrated representation of field sound radiation. Full article

Aerial tiny objects naturally possess conjugate symmetry in the frequency domain, yet complex scenarios, heavy clutter, and frequent rotation/occlusion lead to severe detail loss, inaccurate contour modeling, and high false/miss rates in existing detectors. Current Fourier-based methods neither leverage object symmetry nor coordinate with Fourier analysis to jointly enhance contour representation and spatial-frequency feature learning, suffering from weak fusion, phase-sensitive coefficient regression, and poor discriminability. To fill this gap, we propose the Frequency-Aware Fourier Series Detection Network (FAFSDet), which explicitly exploits the inherent symmetry of tiny objects and their frequency-domain representations. Specifically, FAFC (Frequency-Aware Feature Fusion) employs conjugate-symmetry-guided dynamic low-pass filtering, similarity-based rearrangement, and adaptive high-frequency enhancement to recover degraded symmetric patterns. FSPRM (Fourier Series Profile Representation) utilizes the symmetric positive–negative frequency distribution to achieve compact parametric contour encoding and normalized centroid-shape description. FSDIM (Fourier Series Detection Inference) incorporates symmetric multi-scale branches, a rolling-optimization loss that eliminates phase interference while preserving coefficient-regression symmetry, and inverse Fourier transform for precise contour reconstruction and end-to-end detection. Extensive experiments on DOTA, AI-TOD, and UCAS-AOD demonstrate that our method achieves superior performance (mAP 82.18%, 51.2%, and 90.70%, respectively) and strong generalization, particularly in scenarios where symmetry is most severely compromised, confirming that exploiting these symmetry properties substantially boosts detection accuracy. Full article