Search Results (644)

This study evaluates how architectural heterogeneity and reward design influence coordination in decentralized IPPO-based Multi-Agent Reinforcement Learning (MARL) systems operating in Search-and-Rescue (SAR) environments with strict sequential task dependencies. Controlled experiments compare homogeneous and heterogeneous teams across multiple reward structures, observation ranges, task complexities, and asymmetric role distributions in a partially observable, communication-denied, decentralized environment. The results show that homogeneous teams consistently achieve faster convergence, higher coordination efficiency, and more stable cooperative behavior than balanced heterogeneous configurations. Under hybrid global–local reward schemes, homogeneous teams achieve a higher success rate than local-only reward schemes. A dedicated test confirms this advantage stems from team composition itself, not parameter sharing: it persists even under independent, non-shared policy networks. Additional experiments suggest that asymmetric role allocation may influence coordination dynamics in heterogeneous teams. The study provides empirically grounded design recommendations for cooperative multi-agent systems, derived from controlled simulation experiments under communication-limited sequential coordination constraints. Full article

We study the unsteady mechanical response of an incompressible viscoelastic polymeric fluid in a flat channel, governed by the Vinogradov–Pokrovskii rheological model. The motion arises from an electrohydrodynamic reduction of Poiseuille type, after which the mechanical subsystem decouples from the electric field; the velocity then depends on time and on the transverse coordinate only. Treating the rheological parameter as small, we reduce the governing system in the leading-order approximation to a non-autonomous second-order evolution equation whose stiffness coefficient relaxes exponentially in time, so that the nonstationarity is driven by the internal relaxation of the normal stress rather than by an external force. For spatially homogeneous initial normal stress, we diagonalize the Galerkin system in the sine basis and obtain an explicit modal representation in which each mode satisfies a Bessel equation whose order depends on the mode number. This yields a critical index that splits the modes into three regimes—real order, zero order, and purely imaginary order—a structure absent from the classical UCM and Oldroyd-B solutions. Using the explicit representation, we prove convergence of the modal series and show that the solution decays in the long-time limit, so that the rest state is asymptotically stable in the natural energy phase space. The analytical solution is confirmed numerically. Full article

The generalized Burgers–Huxley (gBH) equation arises as a canonical model in nerve-pulse propagation (generalizing the Hodgkin–Huxley/FitzHugh–Nagumo excitable-media framework), in population dynamics with Allee-threshold reaction kinetics, and in nonlinear wave propagation in dispersive media; accurate benchmark solutions are essential for quantitative predictions in these domains. We couple the fourth-order exponential time differencing scheme ETDRK4 with a Chebyshev collocation spatial discretization and a linear boundary-lifting procedure to solve the gBH equation on a bounded interval with non-homogeneous Dirichlet data. On the canonical Ismail–Raslan–Rabboh travelling-wave benchmark the scheme attains $L^{\infty }$ errors at the level of floating-point round-off (∼10⁻¹⁹ absolute, ∼10⁻¹⁵ relative) with as few as $N=2$ collocation points and a single time step of size $\Delta t=1.0$—that is, three total nodes and one ETDRK4 advance. In strongly nonlinear regimes ($\gamma =0.1, 0.3, 0.5, 0.9$) the scheme exhibits approximately $\mathcal{O}(\Delta t^{2.45})$ temporal convergence across all four parameter values, consistent with the classical Hochbruck–Ostermann order reduction for exponential integrators on parabolic PDEs with non-homogeneous Dirichlet data. Used as a high-accuracy probe, the scheme provides a diagnostic of independent interest: the wave-speed formula of Wang, Zhu and Lu, still appearing as the exact-solution benchmark in numerical studies as recently as 2020, does not satisfy the partial differential equation. The corrected formula stated by Deng and verified symbolically by Appadu and Tijani is the unique value that makes the travelling-wave ansatz a genuine solution. We derive the residual associated with Wang’s formula in closed form, $R=\frac{\gamma A_1}{2}(A_2-A_2^{\mathrm{W}})(1-v^2)$, and show both analytically and numerically that reported errors for schemes benchmarked against Wang’s formula coincide with the analytical wave-profile gap $\frac{\gamma A_1}{2}|A_2-A_2^{\mathrm{W}}|$ rather than with true scheme accuracy. At the Ismail benchmark this gap equals $3.748\times {10}^{-7}$, which matches the N- and $\Delta t$-independent plateau observed when the scheme is measured against Wang’s profile. In the nerve-pulse and excitable-media interpretation, the two formulas correspond to action-potential propagation speeds of opposite sign at the Ismail benchmark, underscoring that the correction is not a mere algebraic curiosity but changes the qualitative physical prediction of the model. Full article

The solid-phase volume fraction is a core process parameter that determines the mixing quality of high-viscosity polyurethane adhesives. It exhibits complex nonlinear couplings with mixing homogeneity, rheological properties, and energy consumption. Traditional trial-and-error experiments and standalone CFD simulations are hindered by high costs, long computational cycles, and inefficient parameter optimization processes. Consequently, these limitations prevent them from satisfying industrial-scale process optimization demands. To address these challenges, this study proposes a single-variable, multi-criteria intelligent optimization method integrating CFD with machine learning. This study uses a 5000 L industrial stirred reactor as the research object. Twelve typical operating conditions were designed within the 8–32% solid-phase volume fraction range. Numerical simulations were conducted based on the laminar Mixture multiphase flow model. These simulations established a small-sample, high-fidelity operating condition database. To mitigate underfitting caused by limited CFD data, an adaptive cubic spline interpolation algorithm was employed for dataset augmentation. Subsequently, a comparative analysis was conducted between Gaussian process regression (GPR) and support vector regression (SVR) as small-sample surrogate models. Test results demonstrate that the GPR model achieves a coefficient of determination (R²) of 0.968, with significantly superior prediction accuracy and generalization performance compared to the SVR model. This study uses the solid-phase volume fraction as the sole decision variable. A five-criteria optimization model was constructed based on the optimal GPR surrogate, combined with the NSGA-II algorithm and ideal point decision criterion. Three optimal operating modes were identified—balanced, low-energy, and high-dispersion—tailored for different production scenarios. The proposed integrated collaborative optimization approach effectively reduces process development costs. It also provides comprehensive theoretical foundations and technical references for intelligent control and energy-efficient production of high-viscosity non-Newtonian solid–liquid mixing systems. Full article

Mechanical anisotropy plays a crucial role in structural engineering. Conventional periodic lattices offer limited control over anisotropy, largely due to tessellation constraints. Spinodal metamaterials, inspired by the eponymous phase-separation process, provide non-symmetric but periodic and tunable architectures capable of exhibiting strong directional stiffness. In this work, both isotropic and anisotropic spinodal decomposition are modeled, with anisotropy introduced through directional mobility coefficients along the principal spatial axes. The resulting phase fields are converted into STL models and analyzed via finite element-based homogenization to compute directional Young’s moduli. A genetic algorithm identifies parameter combinations that maximize stiffness along a target direction while minimizing it along orthogonal directions. The results demonstrate a direct correspondence between the anisotropy of the spinodal transformation and the mechanical response of the resulting metamaterial, thereby highlighting the advantages of modeling an anisotropic diffusional process, albeit with a greater computational cost. Full article

Federated learning (FL) enables privacy-preserving fault diagnosis across distributed industrial devices, but most existing methods assume homogeneous model architectures and comparable client resources. This assumption is unrealistic in resource-constrained Industrial Internet of Things (IIoT) scenarios, where clients may have substantially different memory and computation capacities. To address this challenge, we propose FedAnchor, an anchored and adaptive FL framework for resource-heterogeneous fault diagnosis. FedAnchor decomposes each client submodel into a shared anchored core and a client-specific adaptive extension. The anchored core provides a common parameter subspace for consistent masked aggregation, while the adaptive extension is selected by a server-side reinforcement-guided policy under client memory budgets. This design couples resource-aware submodel allocation with structurally aligned aggregation. Experiments on four benchmark datasets and six heterogeneous memory configurations show that FedAnchor achieves competitive or superior accuracy compared with representative homogeneous and model-heterogeneous FL baselines. Under the evaluated non-IID settings, FedAnchor improves accuracy by up to 9.8 percentage points over the strongest baseline, while maintaining favorable communication–accuracy trade-offs and empirical stability. Full article

Hydrogen sampling is an essential part of ensuring reliable and accurate hydrogen fuel quality for expanding heavy-duty vehicle applications. Hydrogen sampling at refuelling stations is highly sensitive to operational conditions, especially temperature, pressure, storage homogeneity, and nozzle-purging procedures. The direct sampling method operates with a hydrogen fuelling station in maintenance mode and requires that parameters be set properly. This study investigated the impact of temperature, pressure, storage bank selection, and venting on hydrogen sample quality. This study shows that hydrogen sampling at refuelling stations is strongly influenced by operational parameters, with temperature and pressure mainly affecting the water content while other contaminants remain largely stable; storage bank composition and insufficient nozzle purging can also significantly bias results through contamination or non-representative sampling. To ensure reliable measurements, this study recommends conducting sampling under representative operational conditions, including matching the delivery temperature and nominal delivery pressure, verifying storage homogeneity, and applying adequate nozzle-purging procedures. However, further validation across different systems is still needed. Full article

To reveal the influence of pulsed current density on the microstructural evolution and mechanical properties of Mg-Gd-Y-Zn-Zr rare-earth magnesium alloy wires, extruded Mg-10Gd-3.4Y-1.3Zn-0.4Zr alloy wire was selected as the research material. By regulating the current density in the range of 12–20 A/mm², the effects on temperature rise behavior, microstructural evolution, and mechanical properties were systematically investigated. The results show that as the current density increases from 12 A/mm² to 20 A/mm², the measured surface peak temperature rises from 207 °C to 497 °C, and the mechanical properties among the electropulsing-treated samples exhibit a trend of first increasing and then decreasing. Among these treated samples, the optimal combination of strength and ductility is achieved at a current density of 15 A/mm², at which the tensile strength and elongation reach 312.2 MPa and 13.6%, respectively. Microstructural analysis indicates that appropriate pulsed electrical parameters promote the dissolution, fragmentation, and homogenized dispersion of block-shaped long-period stacking ordered (LPSO) phases, thereby optimizing the internal strain state and facilitating the activation of non-basal <c+a> slip. However, when the current density increases to 20 A/mm², excessive thermal input leads to grain coarsening and a network-like W-phase precipitation, indicating that excessive energy input can lead to microstructural instability and mechanical degradation. Full article

This study develops a stochastic queueing framework to model the dynamical interaction between proliferating tumor cells and the adaptive immune system, incorporating the critical phenomenon of immune checkpoint inhibition. The immune response is conceptualized as a multi-server service system where CD8+ T cells and natural killer cells act as parallel servers eliminating cancerous cells, while tumor cell division follows a controlled birth process with logistic growth constraints. The novelty lies in embedding a state-dependent immune evasion mechanism through a dimensionless checkpoint parameter $\kappa \in [0,1]$ that modulates the effective killing rate as a function of tumor burden via the saturation function $\varphi \left(n\right)=n/(K+n)$. The resulting process $\left\{N\right(t),t\ge 0\}$ constitutes a non-homogeneous birth–death Markov chain with the following transition rates: $\begin{array}{ccc}\hfill {\lambda }_n& =\lambda \left(1-{\displaystyle \frac{n}{K}}\right){\mathbb{1}}_{\{n<K\}},{\mu }_n\hfill & \hfill =\left\{\begin{array}{cc}n\mu (1-\kappa \varphi (n\left)\right),\hfill & 1\le n<c,\hfill \\ c\mu (1-\kappa \varphi (n\left)\right),\hfill & n\ge c,\hfill \end{array}\right.\end{array}$ We derive the exact stationary distribution in closed form, establishing the fundamental stability condition ${\rho }_{\mathrm{eff}}=\lambda /\left[c\mu (1-\kappa )\right]<1$ for tumor elimination. The expected tumor burden L and mean elimination time W are computed explicitly. A critical threshold ${\kappa }_c=1-\lambda /\left(c\mu \right)$ emerges, above which the immune system loses control regardless of other parameters. Heavy-traffic analysis reveals quadratic divergence $L\sim K/{(1-{\rho }_{\mathrm{eff}})}^2$ as ${\rho }_{\mathrm{eff}}\to 1^{-}$, explaining catastrophic tumor escape. Extensive Monte Carlo simulations ($n={10}^4$ replications) validate theoretical predictions with relative errors <5% and p-values $>0.05$ from two-sample t-tests. The model provides quantitative tools for optimizing checkpoint inhibitor dosages and predicting patient-specific responses in immuno-oncology. Full article

The present research investigated the effects of two pork back-fat concentrations (30% fat, FAT30, and 40% fat, FAT40), two sodium chloride levels (2.4% and 2.6%), and two starter culture combinations (Lactobacillus curvatus/Staphylococcus xylosus; LAB6, and Lactobacillus sakei/Staphylococcus xylosus; LAB8) on ripened ostrich salami. Salami samples were formulated without nitrite and nitrate, which aligns with consumer demands for healthier, cleaner-label meat products. It is specified that the present experiment is structured with a single-batch-per-treatment combination: this was due to structural processing limitations in the production facility, which was an artisanal laboratory and not an industry plant. After 10 weeks of ripening, FAT30 salami showed higher values of pH, salt content, water-phase salt (WPS), α-tocopherol, free fatty acids (FFA), and secondary lipid oxidation products (TBARS) compared with FAT40 salami. Conversely, FAT40 salami exhibited higher water activity (a_w), moisture-to-protein ratio (M:P), conjugated dienes (CD; primary lipid oxidation products), and non-protein nitrogen (NPN) than FAT30 salami. Both NaCl concentration and starter culture type influenced several of the measured variables. Specifically, salami containing 2.4% salt exhibited higher FFA and CD values than the formulation containing 2.6% salt. Likewise, the LAB8 starter culture resulted in higher CD and NPN levels compared with LAB6. Fat inclusion level significantly affected sensory characteristics. FAT40 salami exhibited greater intensities of gamy, metallic, fatty, and moldy flavors, as well as higher overall off-flavor intensity, tenderness, and juiciness. In contrast, FAT30 salami was characterized by greater cohesiveness and a more pronounced ripening flavor. The 2.6% sodium chloride treatment resulted in greater color homogeneity, higher odor intensity, and stronger rancid notes, while reducing the perception of metallic, fatty, and moldy flavors compared with the 2.4% treatment. Salami inoculated with LAB6 exhibited a higher intensity of off-flavors than the formulation produced with LAB8. Moreover, several significant interactions among the three experimental factors were observed. After 20 weeks of ripening, the effects observed after 10 weeks for most physicochemical parameters were largely maintained. However, FFA and CD concentrations (both below the limit of quantification) no longer differed between the two fat inclusion levels. Sensory evaluation revealed that the differences between FAT30 and FAT40 in undesirable flavor attributes disappeared over time, whereas the perception of ripening and maturity became even more pronounced in FAT30 salami. Regarding FA composition, FAT30 salami contained higher proportions of saturated FA and polyunsaturated FA, whereas FAT40 salami was characterized by a higher monounsaturated FA content and more favorable lipid quality indices. Full article

Background: Skin aging is influenced by intrinsic and extrinsic factors that alter skin texture and microrelief. Non-invasive imaging techniques enable objective longitudinal assessment of these characteristics. This study aimed to instrumentally and subjectively evaluate skin surface changes during a standardized 90-day dermocosmetic regimen containing multiple active ingredients. Methods: This prospective, single-arm observational study included 47 women aged 30–60 years with visible signs of facial skin aging. Participants applied a standardized anti-age serum and cream regimen twice daily for 90 days. Skin surface morphology was assessed at baseline and after 30, 60, and 90 days using Visioscan^® VC 20plus imaging with SELS^® software. Subjective outcomes were evaluated using a structured product questionnaire and the validated Skindex-16 instrument. Results: Significant temporal changes were observed across objective skin surface parameters. From baseline to day 90, mean skin roughness decreased from 4.99 ± 2.29 to 1.66 ± 0.67, while the wrinkle-related parameter decreased from 191.08 ± 87.26 to 74.27 ± 15.25. Reductions were also observed in scaliness, contrast, entropy, variance, and anisotropy, whereas smoothness, homogeneity, and energy increased over time. Skindex-16 scores decreased across Symptoms, Emotions, and Functioning domains, indicating lower participant-reported dermatology-related burden. The regimen was well tolerated, with no serious adverse events reported. Conclusions: In this exploratory single-arm observational study, the standardized dermocosmetic regimen was associated with measurable longitudinal changes in Visioscan-derived skin surface parameters and patient-reported outcomes over 90 days. Because the study included no control group, the findings cannot establish causality or isolate the contribution of individual regimen components. Full article

Existing target sorting methods suffer severe performance degradation or even failure under inherent severe spectrum overlap, homogeneous protocol parameters, and scarce single-source points in Synchronous Non-Orthogonal Frequency Hopping (SNOFH) scenarios. To address this challenge, this paper proposes a passive sorting framework for SNOFH UAV swarm signals based on multi-station relative hopping time difference. The proposed framework constructs a spatial-location-driven sorting feature system, designs a kernel joint distribution adaptation module to eliminate inter-station measurement discrepancies, and develops a multi-scale wavelet-based method to achieve sub-sampling level hopping time extraction, reducing the dependence on prior FH parameters and hardware radio frequency fingerprints. Experimental comparisons between the proposed and reference sorting methods are conducted on a simulated SNOFH dataset to validate the performance of the proposed sorting framework. The experimental results show that the proposed method achieves the highest sorting accuracy of 98%, outperforming adopted baselines in most SNOFH cases. The proposed method exhibits favorable robustness with noise interference, clock-synchronization error, carrier-frequency offset and multipath influence. It is a suitable choice for UAV swarm sorting under regular and slow-varying UAV formations. Full article

Spirulina-derived protein hydrolysates (SPPHs) have attracted considerable attention as bioactive agents due to their potential metabolic and physiological benefits. This study evaluated the therapeutic efficacy of different enzyme-specific SPPHs—Pepsin (SPPH-P), Trypsin (SPPH-T), Chymotrypsin (SPPH-C), and a combined hydrolysate (SPPH-PTC)—in high-fat diet (HFD)-induced male Wister rats, compared with Spirulina platensis protein extract (SPPE, formulated using freeze–thaw cycles and ultrasonication followed by centrifugation) and atorvastatin as a Positive Control. The animals were randomly allocated into seven groups (n = 6 per group) and received their respective treatments orally for 4 weeks. Across treatment groups, significant improvements in obesity-related anthropometric indices were observed, including reductions in BMI, Lee Index, and abdominal circumference to thoracic circumference ratio (AC:TC), with the strongest effects noted in the atorvastatin and SPPH-PTC groups. Protein metabolism markers showed enhanced hepatic and serum protein status, reflected by increased albumin and total protein concentrations. Lipid profile analysis revealed marked decreases in total cholesterol, triglycerides, and LDL in both serum and liver homogenates, while HDL exhibited non-significant but favorable elevations. Liver function markers (bilirubin, ALT, AST) and renal parameters (uric acid, BUN) demonstrated notable improvements, particularly in enzyme-derived hydrolysate groups and Positive Control. Antioxidant assessments indicated substantial reductions in MDA levels and significant increases in SOD, CAT, and GSH activities in serum and liver tissues, confirming enhanced oxidative stress resistance. Among all treatments, SPPH-PTC consistently produced the most robust therapeutic outcomes. Overall, Spirulina protein hydrolysates, especially the combined PTC formulation, exert comprehensive beneficial effects on metabolic regulation, hepatic and renal function, and oxidative balance. These findings support their potential application as functional bioactive agents for managing obesity-associated metabolic disturbances. Full article

ZERODUR glass-ceramic is widely used in ultra-precision optical components because of its extremely low thermal expansion and excellent dimensional stability. However, its two-phase microstructure, composed of crystalline and amorphous phases with different mechanical properties, may cause non-uniform material removal during magnetorheological polishing, thereby limiting further improvement of nanoscale surface quality. To address this issue, this study investigates the effect of oxide abrasives on the surface homogenization of ZERODUR. A single-particle abrasive–workpiece contact model based on modified Hertz contact theory and elastoplastic contact analysis was established to compare the indentation responses of CeO₂, SiO₂, and ZrO₂ abrasives in the two constituent phases. Magnetorheological polishing experiments were conducted under identical process parameters, and the polished surfaces were characterized by AFM over scan areas of 2 μm × 2 μm, 5 μm × 5 μm, and 10 μm × 10 μm. The results show that all three abrasives improved the surface quality of the ring-polished substrate, with ZrO₂ achieving the best surface homogenization performance. The lowest roughness, Ra = 0.104 nm, was obtained at a 2 μm field of view, and the ZrO₂-polished surface showed more stable roughness evolution across different scan sizes than the CeO₂- and SiO₂-polished surfaces. These results indicate that the elastic modulus, hardness, and mechanical compatibility of abrasives with ZERODUR play key roles in governing contact stress, indentation behavior, and final surface quality. This work addresses the lack of mechanistic understanding of abrasive-dependent surface homogenization in the magnetorheological polishing of two-phase ZERODUR glass-ceramic. The main innovation is the integration of contact-mechanics-based abrasive–workpiece modeling with multi-scale AFM characterization to clarify how abrasive mechanical compatibility affects nanoscale surface uniformity and to guide abrasive selection for ultra-smooth optical manufacturing. Full article

Pavement texture strongly affects skid resistance, drainage, and tire–pavement contact stability, yet its transverse evolution under wheel-track-concentrated loading remains insufficiently quantified. This study proposes a 3D-laser-scanning-based framework for evaluating the transverse homogeneity of preventive maintenance pavements during service. Ten field sections on the Jiangluo Expressway in Guangdong Province, China, covering five preventive maintenance surface systems at two service stages (six months and one year), were investigated. Reflection intensity histogram features and geometric texture parameters were screened against transverse wheel-track distribution to identify representative indicators of asphalt film peeling and aggregate wear. Weighted average grayscale was selected as the optical indicator, whereas height-distribution kurtosis was selected as the geometric indicator. A section-level homogeneity index based on normalized median absolute deviation was then used to quantify transverse dispersion. The results show that weighted average grayscale and kurtosis are the most sensitive of the tested indicators to transverse wheel-track distribution, with $R^2=0.973$ and $R=0.9057$, respectively. Wheel-track regions generally exhibited more severe optical and geometric deterioration than non-wheel-track regions, and transverse homogeneity tended to decrease from six months to one year. Within the investigated expressway sections, the framework was sensitive to different degrees of service-stage transverse wear evolution; however, broader multi-site validation is still required before threshold-based general applications can be established. Full article