2585 papers
This study addresses global subnational sanitation data gaps in low- and middle-income countries by applying a Tabular Prior-Fitted Networks model to survey data from 1.8 million households, revealing stark inequalities in coverage and highlighting the severe lack of access and high uncertainty in fragile, low-income sub-Saharan African settings.
This study numerically optimizes the air inlet structure of an H₂/CH₄ micromix combustor nozzle by evaluating hole arrangement and diameter grading, demonstrating that multi-stage vortex generation and redistributed air momentum significantly enhance mixing uniformity and fuel flexibility across a wide range of hydrogen blending ratios.
This paper introduces a radiometrically-agnostic SegFormer framework that fuses C-band SAR imagery with topographic data to achieve state-of-the-art, efficient water segmentation across diverse sensors, demonstrating that a lightweight B0 model outperforms its larger counterpart in both accuracy and inference speed.
This paper presents the design and experimental evaluation of a bifacial solar-powered EV charging system, demonstrating that optimizing rear-side reflectivity with aluminum foil significantly enhances energy generation by 15–25% and improves charging stability compared to traditional monofacial systems.
This paper proposes a unified continuous-time framework for adaptive smartphone battery prediction that integrates hardware dynamics, temperature effects, and aging models, demonstrating significant error reduction through load calibration and recursive comparison while identifying specific conditions for model deployment and the necessity of complete-discharge validation.
This study validates a meshless collocation framework for analyzing the free vibration of multilayer functionally graded hyperelastic beams through experimental testing and finite element comparison, while also demonstrating the effectiveness of an artificial neural network in accurately predicting the structure's natural frequencies.
This paper presents an integrated framework that shifts seismic risk assessment from focusing on direct physical damage to quantifying system-level societal impacts across interdependent infrastructure, revealing that prioritizing bridge retrofitting to maintain hospital access is often more effective for resilient urban planning than traditional damage-based approaches.
This paper presents a portable, multimodal screening platform that integrates gravimetric–ultrasonic density measurement and smartphone-based colorimetry to preliminarily assess the authenticity of patchouli oil, demonstrating a strong correlation with reference values (R² = 0.9749) while emphasizing its role as a non-standardized screening tool rather than a substitute for laboratory authentication.
This paper presents a computationally efficient design optimization framework for modular robotic arms that utilizes a genetic algorithm-trained neural network surrogate to replace expensive finite-element analyses, successfully achieving a 19.9% mass reduction while satisfying all structural constraints with significantly fewer high-fidelity evaluations.
This study utilizes a novel quasi-dynamic test rig to simulate impeller wake migration in a high-flow-coefficient centrifugal compressor, revealing that while the U-bend significantly reduces circumferential non-uniformity and return channel vanes induce new fluctuations, the total stage loss remains constant regardless of the wake's circumferential position.