Search Results (2,061)

Retrieval-Augmented Generation (RAG) is widely used for long-text summarization due to its efficiency and scalability. However, standard RAG methods flatten documents into independent chunks, disrupting sequential flow and thematic structure, resulting in significant loss of contextual information. This paper presents MOEGAT, a novel graph-enhanced retrieval framework that addresses this limitation by explicitly modeling document structure. MOEGAT constructs an Orthogonal Context Graph to capture sequential discourse and global semantic relationships—long-range dependencies between non-adjacent text spans that reflect topical similarity and logical associations beyond local context. It then employs a query-aware Mixture-of-Experts Graph Attention Network to dynamically activate specialized reasoning pathways. Experiments conducted on three public long-text summarization datasets demonstrate that MOEGAT achieves state-of-the-art performance. Notably, on the WCEP dataset, it outperforms the previous state-of-the-art Graph of Records (GOR) baseline by 14.9%, 18.1%, and 18.4% on ROUGE-L, ROUGE-1, and ROUGE-2, respectively. These substantial gains, especially the 14.9% improvement in ROUGE-L, reflect significantly better capture of long-range coherence and thematic integrity in summaries. Ablation studies confirm the effectiveness of the orthogonal graph and Mixture-of-Experts components. Overall, this work introduces a novel structure-aware approach to RAG that explicitly models and leverages document structure through an orthogonal graph representation and query-aware Mixture-of-Experts reasoning. Full article

Spontaneous coal combustion accounts for more than 90% of mine fires, and at the same time, the ‘dual carbon’ strategy requires fire prevention and extinguishing materials to have both low-carbon and environmentally friendly functions. To meet on-site application needs, a composite gel with fast injection, flame retardant, and CO₂ adsorption functions was developed. PVA-PEI-PAC materials were selected as the gel raw materials, and an orthogonal test with three factors and three levels was used to optimize the gelation time parameters to identify the optimal formulation. The microstructure of the gel, CO₂ adsorption performance, as well as its inhibition rate of CO, a marker gas of coal spontaneous combustion, and its effect on activation energy were systematically characterized through SEM, isothermal/temperature-programmed/cyclic adsorption experiments, and temperature-programmed gas chromatography. The results show that the optimal gel formulation is 14% PVA, 7% PEI, and 5.5% PAC. The gel microstructure is continuous, dense, and rich in pores, with a CO₂ adsorption capacity at 30 °C and atmospheric pressure of 0.86 cm³/g, maintaining over 76% efficiency after five cycles. Compared with raw coal, a 10% gel addition reduces CO release at 170 °C by 25.97%, and the temperature-programmed experiment shows an average CO inhibition rate of 25% throughout, with apparent activation energy increased by 14.96%. The gel prepared exhibited controllable gelation time, can deeply encapsulate coal, and can efficiently adsorb CO₂, significantly raising the coal–oxygen reaction energy barrier, providing an integrated technical solution for mine fire prevention and extinguishing with both safety and carbon reduction functions. Full article

On-orbit cutting is a critical process for the on-orbit manufacturing of carbon fiber reinforced polyetheretherketone composites (CF/PEEK) truss structures, with pulsed laser cutting serving as one of the feasible methods. Achieving high-quality cutting of CF/PEEK remains a major challenge for on-orbit manufacturing. Therefore, the cutting process of CF/PEEK prepreg tape was studied by an ultraviolet (UV) nanosecond laser. A three-factor, five-level orthogonal experiment was carried out to analyze the influence of laser repetition rate (LRR), laser cutting speed (LCS), and laser scanning times (LCTs) on cutting quality. The ablation mechanism dominated by the photothermal effect between the UV nanosecond laser and CF/PEEK was analyzed, and the by-products in the cutting process were explored. Finally, the optimal cutting quality (the width of slit (W_s) = 41.69 ± 3.54 μm, the heat-affected zone (HAZ) = 87.27 ± 7.30 μm) was obtained under the process conditions of LRR 50 kHz-LCS 50 mm/s-LCT 16 times. The findings show that the W_S and HAZ increase with the increase in LRR and LCT and the decrease in LCS, and the carbon fiber decomposes and escapes due to the photothermal effect. Full article

Canagliflozin (CFZ), a sodium–glucose cotransporter 2 (SGLT2) inhibitor, is extensively utilized in the management of type 2 diabetes. Among its various polymorphic forms, the hemi-hydrate (Hemi-CFZ) has been selected as the active pharmaceutical ingredient (API) for CFZ tablets due to its superior solubility. However, during the production, storage, and transportation of CFZ tablets, Hemi-CFZ can undergo transformations into anhydrous (An-CFZ) and monohydrate (Mono-CFZ) forms under the influence of environmental factors such as temperature, humidity, and pressure, which may adversely impact the bioavailability and clinical efficacy of CFZ tablets. Therefore, it is imperative to develop rapid, accurate, non-destructive, and non-contact methods for quantifying An-CFZ and Mono-CFZ content in CFZ tablets to control polymorphic impurity levels and ensure product quality. This research evaluated the feasibility and reliability of using near-infrared spectroscopy (NIR) combined with partial least squares regression (PLSR) for simultaneous quantitative analysis of An-CFZ and Mono-CFZ in CFZ tablets, elucidating the quantifying mechanisms of the quantitative analysis model. Orthogonal experiments were designed to investigate the effects of different pretreatment methods and ant colony optimization (ACO) algorithms on the performance of quantitative models. An optimal PLSR model for simultaneous quantification of An-CFZ and Mono-CFZ in CFZ tablets was established and validated over a concentration range of 0.0000 to 10.0000 w/w%. The resulting model, Y_{An-CFZ/Mono-CFZ} = 0.0207 + 0.9919 X, achieved an R² value of 0.9919. By analyzing the relationship between the NIR spectral signals selected by the ACO algorithm and the molecular structure information of An-CFZ and Mono-CFZ, we demonstrated the feasibility and reliability of the NIR-PLSR approach for quantifying these polymorphic forms. Additionally, the mechanism of PLSR quantitative analysis was further explained through the variance contribution rates of latent variables (LVs), the correlations between LVs loadings and tablets composition, and the relationships between LV scores and An-CFZ/Mono-CFZ content. This study not only provides a robust method and theoretical foundation for monitoring An-CFZ and Mono-CFZ content in CFZ tablets throughout production, processing, storage, and transportation, but also offers a reliable methodological reference for the simultaneous quantitative analysis and quality control of multiple polymorphic impurities in other similar drugs. Full article

The use of large-aperture scanning mirrors for high-resolution and wide-swath imaging represents a major trend in Earth observation technology. However, to improve dynamic response performance, scanning mirror assemblies are highly lightweighted, resulting in reduced overall stiffness. This makes the mirror surface susceptible to thermal and inertial loads during operation, leading to degraded surface accuracy and poor imaging quality. Moreover, dynamic scanning mirror has the multi-disciplinary coupling effects and non-circular structural characteristics. It poses significant challenges for surface figure analysis. To address these issues, this paper proposes a surface analysis method for high-dynamic scanning mirrors under multi-physics coupling in non-circular domains. First, a finite element model of the mirror assembly is established based on the minimum aperture and angular velocity parameters. Through finite element analysis, the surface response of the scanning mirror assembly under thermal loads, dynamic inertial loads, and their coupled effects is quantitatively investigated. Subsequently, an analytical approach, which combines rigid-body displacement separation and Gram–Schmidt orthogonalization, is developed to construct non-circular Zernike polynomials, enabling high-precision fitting and reconstruction of the mirror’s dynamic surface distortions. Numerical experiments validate the accuracy of the model. Results show that for a scanning mirror with an aperture of 466 mm × 250 mm under the coupled condition of a 5 °C temperature rise and 50 N·mm torque, the surface figure achieves RMS < 2 nm and PV < 22 nm, with a fitting accuracy achieves 10⁻⁶. These results verify the accuracy and reliability of the proposed method. The surface analysis approach presented in this study provides theoretical guidance and a design framework for subsequent image quality evaluation and assurance. Full article

The rising global demand for prosthetic limbs, driven by approximately 185,000 amputations annually in the United States, underscores the need for innovative and cost-efficient solutions. This study explores the integration of hybrid materials, advanced 3D printing techniques, and smart sensing technologies to enhance prosthetic finger production. A Taguchi-based design of experiments (DoE) approach using an L09 orthogonal array was employed to systematically evaluate the effects of infill density, infill pattern, and print speed on the tensile behavior of FDM-printed PLA components. Findings reveal that higher infill densities (90%) and hexagonal patterns significantly enhance yield strength, ultimate tensile strength, and stiffness. Additionally, the rheological properties of polydimethylsiloxane (PDMS) were optimized at various temperatures (30–70 °C), characterizing its viscosity, shear-thinning factors, and stress behaviors for 3D bioprinting of flexible sensors. Barium titanate (BaTiO₃) was incorporated into PDMS to fabricate a flexible tactile sensor, achieving reliable open-circuit voltage readings under applied forces. Structural and functional components of the finger prosthesis were fabricated using FDM, stereolithography (SLA), and extrusion-based bioprinting (EBP) and assembled into a functional prototype. This research demonstrates the feasibility of integrating hybrid materials and advanced printing methodologies to create cost-effective, high-performance prosthetic components with enhanced mechanical properties and embedded sensing capabilities. Full article

To avoid predators, benthic fish will stir up the sediment on the seabed by flapping their pectoral fins, thus burying themselves. This self-burial concealment strategy can offer bionic enlightenment for the benthic residence method of Unmanned Underwater Vehicles (UUVs). In this paper, based on the observation results of the self-burial behavior of benthic fish, a two-dimensional fluid-particle numerical model was developed to simulate the processes of sediment transport induced by pectoral fin flapping. In addition, an orthogonal experimental design was employed to analyze the effects of body length, flapping amplitude, flapping number, flapping frequency, and particle size on burial ratio, input power, and burial efficiency. The results reveal that rapid pectoral fin flapping enables benthic fish to fluidize sediments and achieve self-burial. Among the influencing factors, body size has the most significant impact on coverage ratio and input power, as larger fish generate stronger tip vortices and fluid disturbances, making local flow velocities more likely to exceed the critical starting velocity. In contrast, particle size has the weakest effect on burial performance, while kinematic parameters exert a far greater impact on self-burial than environmental parameters. The research results can offer references for the biomimetic design of self-burying UUVs. Full article

Elastic wavefield decomposition into quasi-compressional (qP) and quasi-shear-vertical (qSV) modes is essential for elastic imaging and inversion in VTI media, but becomes computationally expensive when polarization vectors vary strongly in space. I propose a tensor-train (TT) representation of mixed-domain decomposition projectors, constructed via TT-cross with a single user-specified tolerance and applied efficiently using FFT-based operations. A residual-orthogonal strategy extracts qSV from the residual wavefield after qP removal to suppress mode leakage. The method is implemented in Python/PyTorch with GPU acceleration. Numerical experiments on three 2D VTI models (a two-layer benchmark, a BP 2007 benchmark subset, and an Overthrust-based structurally complex model) demonstrate reconstruction errors of 0.094–0.89% for TT, compared to 1.67–6.44% for a conventional CUR low-rank approach (4–46× improvement), with consistently lower cross-talk and near-unity energy ratios. Time-domain receiver traces further confirm that TT yields smaller reconstruction residual spikes and reduced cross-mode leakage than CUR. Runtime tests show that CUR can be faster on smaller grids, whereas TT with GPU acceleration becomes competitive and can outperform CUR for larger models. The TT representation scales linearly with tensor $\mathcal{O}\left(d N_s r^2\right)$—enabling practical extension to higher-dimensional projector tensors where conven-tional methods become impractical. Full article

This study employed a combined approach of computational fluid dynamics (CFD), numerical simulations, and wind tunnel tests to investigate droplet drift characteristics and develop prediction models in order to address the issues of low pesticide utilization rates and high drift risk, associated with droplet drift during agricultural unmanned aerial vehicle (UAV) spraying, as well as the unreliable results of field experiments. Firstly, a numerical model of the rotor wind field was established using the multiple reference frame (MRF) method, while the realizable k-ε turbulence model was employed to analyze the flow field. The model’s reliability was verified through wind field tests. Next, the Euler–Lagrange method was used to couple the wind field with droplet movement. The drift characteristics of two flat-fan nozzles (FP90-02 and F80-02) were then compared and analyzed. The results showed that the relative error between the simulated and wind tunnel test values was within 20%. Centrifugal nozzle experiments were carried out using single-factor and orthogonal designs to analyze the effects of flight height, rotor wind speed, flight speed, and droplet size on drift. The priority order of influence was found to be “rotor wind speed > flight height > flight speed”, while droplet size (D_V50 = 100–300 µm) was found to have no significant effect. Based on the simulation data, a multiple linear regression drift prediction model was constructed with a goodness of fit R² value of 0.9704. Under the verification condition, the relative error between the predicted and simulated values was approximately 10%. These results can provide a theoretical basis and practical guidance for assessing drift risk and optimizing operational parameters for agricultural UAVs. Full article

Insomnia remains a widespread global health issue, and traditional hypnotic drugs often produce adverse effects. Although spinosin in Ziziphi Spinosae Semen has sleep-promoting effects, its use is limited by poor solubility and low oral bioavailability. In this study, the solvent melt method was used to prepare spinosin solid dispersions, optimising the process with an L₉(3⁴) orthogonal design based on apparent solubility. In vitro dissolution testing showed that solid dispersions of varying particle sizes dissolved more readily than pure spinosin, with smaller particles exhibiting faster dissolution. Cellular uptake was assessed in human colon adenocarcinoma cells, with results revealing enhanced uptake of smaller-particle solid dispersions. Powder X-ray diffraction confirmed that spinosin transformed from a crystalline to an amorphous state in the dispersion system. A quadratic orthogonal experiment was conducted to optimise functional dairy beverage formulation, using the centrifugal sedimentation rate as the evaluation index. In vivo experiments demonstrated that the resulting functional dairy beverage reduced spontaneous activity in mice, achieved a 60% sleep-onset rate, improved ethanol-induced memory impairment and produced marked sleep-promoting effects. Moreover, pharmacokinetic studies confirmed that the spinosin solid-dispersion-based functional dairy beverage significantly enhanced the systemic exposure and oral bioavailability of spinosin compared to the spinosin water suspension. These findings indicate that solid dispersion technology effectively enhances spinosin solubility and that the developed functional dairy beverage shows promise as a sleep-promoting functional food. Full article

This study introduces a hybrid optimization method to enhance the melting characteristics and weld bead morphology of flux-cored wire hardfacing with exothermic addition into the core filler. The Taguchi Design of Experiments (L9 orthogonal array) was used to analyze the effects of key conditions on multiple melting characteristics. The hybrid Taguchi-GRA-PCA effectively identified the best parameter combination, resulting in a significant improvement in overall melting performance. The impact of welding modes on weld bead parameters and melting characteristics was examined. It was determined that the optimal amount of the exothermic addition CuO-Al introduced into the flux-cored wire filler should be a medium level (EA = 28 wt.%). Results showed that wire feed speed WFS and EA had the greatest effect on MOR and DR, while EA and CTWD mainly influenced SF and De. It has been determined that the content of the exothermic additive has a significant impact on the melting process of filler materials, affecting the melting characteristics and weld bead morphology. It has been found that the melting characteristics of deposition rate and spattering factor can be used to optimize welding modes and characterize most output parameters of the welding/surfacing process. Full article

Infrared remote sensing images are often degraded by blur and stripe noise caused by satellite attitude variations, optical distortions, and electronic interference, which significantly compromise image quality and target detection performance. Existing joint deblurring and destriping methods tend to over-smooth image edges and textures, failing to effectively preserve high-frequency details and sometimes misclassifying ringing artifacts as stripes. This paper proposes a variational framework for simultaneous deblurring and destriping of infrared remote sensing images. By leveraging an adaptive structure tensor model, the method exploits the sparsity and directionality of stripe noise, thereby enhancing edge and detail preservation. During blur kernel estimation, a fidelity term orthogonal to the stripe direction is introduced to suppress noise and residual stripes. In the image restoration stage, a WCOB (Non-blind restoration based on Wiener-Cosine composite filtering) model is proposed to effectively mitigate ringing artifacts and visual distortions. The overall optimization problem is efficiently solved using the alternating direction method of multipliers (ADMM). Extensive experiments on real infrared remote sensing datasets demonstrate that the proposed method achieves superior denoising and restoration performance, exhibiting strong robustness and practical applicability. Full article

Incorporating coarse aggregates into ultra-high-performance concrete (UHPC-CA) can reduce material costs, yet reliably predicting its strength-related behavior and overall performance remains challenging. This study examines UHPC-CA through a two-stage orthogonal experimental program comprising 18 mixtures with coarse aggregate, fly ash, and hybrid fiber reinforcements (steel, polypropylene, and composite fibers). Microstructural characterization using scanning electron microscope (SEM) and X-ray computed tomography (X-CT) was conducted to assess interfacial features and crack evolution and to link these observations to the measured mechanical response. Experimentally, fiber reinforcement markedly enhanced post-cracking performance. Compared with the fiber-free control mixture, the optimal hybrid configuration increased flexural strength from 6.9 to 23.5 MPa and compressive strength from 60.1 to 90.5 MPa. The steel–composite fiber system outperformed the steel–polypropylene system, which is consistent with the tighter composite-fiber interfacial bonding observed by SEM/X-CT and supports the feasibility of partially substituting steel fibers. An artificial neural network (ANN) model trained on 50 mixtures and evaluated on 10 unseen mixtures achieved an R² of 0.9703, an MAE of 1.22 MPa, and an RMSE of 2.11 MPa for compressive strength prediction, enabling sensitivity assessment under multi-factor coupling. Overall, the proposed experiment–characterization–modeling framework provides a data-driven basis for performance-oriented mix design and rapid screening of UHPC-CA. Full article

Symmetry and asymmetry jointly govern the dynamics and surface integrity of thin-wall AA7075 end milling. In this work, a symmetry-aware simulation and experimental framework is developed to connect process parameters with milling forces, dynamic response, surface quality, and through-thickness residual stress. A mechanistic milling-force model is first established for multi-tooth end milling, where the periodically repeated tooth-order excitation provides a nominally symmetric load pattern along the tool path. The predicted forces are then used as input for finite-element modal and harmonic-response analysis of a thin-walled component, revealing how symmetric and anti-symmetric mode shapes interact with the tooth-order excitation to generate locally amplified, asymmetric vibration of the compliant wall. Orthogonal and single-factor milling experiments on AA7075 thin-wall specimens are performed to calibrate and validate the force model, and to quantify the influence of feed per tooth, axial depth of cut, spindle speed, and radial width of cut on deformation, surface roughness, and geometric accuracy. Finally, a thermo-mechanically coupled finite-element model is employed to evaluate the residual-stress field, showing a characteristic pattern in which an initially symmetric thermal–mechanical loading produces depth-wise symmetry breaking between tensile surface layers and compressive subsurface zones. The proposed symmetry-aware framework, which combines milling-force theory, finite-element simulation, and systematic experiments, provides practical guidance for selecting parameter windows that suppress vibration, control residual stress, and improve the machining quality of thin-wall AA7075 components. Full article

Electrical discharge machining (EDM) is widely used for machining hard and difficult-to-cut materials; however, the complex and nonlinear nature of the process makes the accurate prediction of key performance indicators challenging, particularly when only limited experimental data are available. In this study, a combined Taguchi design and Gaussian process regression (GPR) modeling framework is proposed to predict the surface roughness (Ra), material removal rate (MRR), and overcut (OC) in die-sinking EDM. An L18 Taguchi orthogonal array was employed to efficiently design experiments involving discharge current, pulse duration, and electrode material. GPR models with an automatic relevance determination (ARD) radial basis function kernel were developed to capture nonlinear relationships and varying parameter relevance. Model performance was evaluated using strict leave-one-out cross-validation (LOOCV). The developed GPR models achieved low prediction errors, with RMSE (MAE) values of 0.54 µm (0.41 µm) for R_a, 1.56 mm³/min (1.21 mm³/min) for MRR, and 0.0065 mm (0.0055 mm) for OC, corresponding to approximately 9.8%, 5.4%, and 5.9% of the respective response ranges. These results confirm stable and reliable predictive accuracy within the investigated parameter domain. Based on the validated surrogate models, multi-objective optimization was performed to identify Pareto-optimal process conditions, revealing graphite electrodes as the dominant choice within the feasible operating region. The proposed approach demonstrates that accurate and robust prediction of EDM performance can be achieved even with compact experimental datasets, providing a practical tool for process analysis and optimization. Full article