The design of adaptive chemotherapy is complicated by nonlinear tumor–immune interactions, delayed drug transport and uncertainty in state evolution. This study develops a delay-aware computational framework that couples a normalized normal–tumor–immune model with central and peripheral pharmacokinetic compartments. A proximal policy optimization (PPO) agent observes the population states, their rates of change and both drug concentrations, and selects a bounded continuous infusion action. Pharmacodynamic killing is driven by peripheral exposure, thereby distinguishing systemic administration from the concentration acting at the tumor site. Positivity and boundedness of the deterministic biological–pharmacokinetic system are established for non-negative initial states and bounded infusion. The learning objective jointly penalizes tumor burden, positive tumor growth, cumulative exposure, abrupt normalized-action variation and depletion of normal and immune populations. Numerical experiments compare untreated and periodic regimens, one- and two-compartment environments, nominal and stochastic transitions, and PPO with two actor–critic baselines. Explicit distribution dynamics change the learned policy from a sustained high-action plateau to an early loading phase followed by gradual tapering. Moderate transition randomization yields the smallest reported tumor-tracking errors and reduced trajectory dispersion, whereas stronger perturbations substantially widen the uncertainty bands. These results identify pharmacokinetic representation as a consequential component of learning-based treatment design. The framework is an in-silico mechanism study rather than a clinical dosing recommendation and requires drug-specific calibration, hard toxicity constraints, complete statistical evaluation and biological validation.
In the digital age, rumor dissemination tends to be diverse and complex. To further analyze its mechanism, this article focuses on the phenomenon of information cocoons, quantifying it as individual opinion intensity and echo chamber intensity. We introduce a general occurrence function to improve the classical SIR rumor propagation model, providing a new perspective for studying new rumor propagation mechanisms. Firstly, a rumor propagation model under dual social platforms is constructed, and the Lyapunov stability theory is used to prove the global asymptotic stability of the equilibrium point. Subsequently, the model is extended to multiplatform scenarios to analyze the global asymptotic stability of propagation dynamics dominated by the basic reproduction number at equilibrium points. To effectively suppress the spread of rumors, a real-time optimization control strategy is implemented to minimize control costs while ensuring that the spread of rumors is contained within the expected time. Finally, numerical simulations are given to validate the effectiveness of the model and explore the impact of different factors on propagation. Further research has shown that the lower the intensity of user opinions and the stronger the echo chamber effect, the slower the disappearance of rumor spreaders and the higher the difficulty of platform control. This article provides theoretical basis and control ideas for rumor management.
With the rapid growth of industrialization, significant changes in the composition of ambient air have been observed, leading to increasing concerns about air pollution. The concentration of air pollutants is commonly represented using a statistical measure known as the Air Quality Index (AQI), which indicates the level of air pollution and its potential adverse effects on human health. In this study, AQI prediction for effective air pollution monitoring in Durgapur city is investigated. The primary objective of this work is to develop an efficient asynchronous cellular automata (ACA)-based model for classifying air pollutant patterns and predicting AQI under both null and periodic boundary conditions. Experimental results demonstrate that the proposed ACA-based approach performs comparably to existing machine learning-based methods for AQI prediction. Moreover, the proposed ACA-based classifier achieves maximum testing accuracies of 88.11% under the null boundary configuration and 91.83% under the periodic boundary configuration, outperforming several existing classification approaches.
Reversibility is a fundamental property of cellular automata (CA) and plays an important role in applications such as cryptography, image security, and pseudo-random number generation. In this paper, we investigate the reversibility of one-dimensional (1D) linear cellular automata (LCA) with radius 2 under reflective boundary conditions (RBC) over a finite field [Formula: see text]. By representing the local rule with a reflective pentadiagonal rule matrix, the reversibility problem is transformed into determining whether the determinant of the corresponding rule matrix is nonzero. We derive a determinant decomposition formula and establish a sixth-order homogeneous linear recurrence relation satisfied by the determinant sequence. Based on this recurrence relation, a fast binary exponentiation (FBE) algorithm is proposed to determine the reversibility of large-scale LCA, reducing the computational complexity from [Formula: see text] of conventional Gaussian elimination to [Formula: see text] with constant space complexity. Furthermore, exploiting the periodicity of linear recurrence sequences over finite fields, a periodicity scale-reduction (PSR) algorithm is developed, which further reduces the complexity to [Formula: see text], where [Formula: see text] is the period determined by the characteristic polynomial of the recurrence. Numerical experiments over [Formula: see text] demonstrate the effectiveness of the proposed methods and reveal the periodic characteristics of reversibility for different local rules. The proposed framework provides an efficient algebraic approach for analyzing the reversibility of large-scale one-dimensional LCA under reflective boundary conditions.
This paper is devoted to the analytical investigation of fractional nonlinear dispersive equations arising in mathematical physics, namely the fractional modified Camassa–Holm (mCH) and modified Degasperis–Procesi (mDP) models. We develop a hybrid transform–iterative framework based on the Aboodh transform combined with a nonlinear recursive scheme to construct explicit series solutions without linearization. A rigorous convergence analysis is established in a Hilbert space setting, where sufficient conditions ensuring the convergence of the iterative sequence are derived using contraction-type arguments. Error estimates for truncated series solutions are also obtained, providing quantitative bounds on the approximation accuracy. The proposed method is applied to several representative problems with different initial profiles, including smooth and non-smooth data, highlighting its robustness in capturing the nonlocal and memory effects induced by fractional operators. The obtained solutions are shown to converge rapidly to the exact solutions, and numerical comparisons demonstrate improved accuracy relative to existing semi-analytical techniques such as the Adomian decomposition method and the homotopy perturbation method. These results indicate that the proposed framework offers an efficient and reliable analytical tool for studying nonlinear fractional evolution equations, with potential applications in the analysis of dispersive wave phenomena in mathematical physics.
Accurate and efficient prediction of hemodynamic parameters is essential for assessing the rupture risk of cerebral aneurysms, particularly in the middle cerebral artery (MCA). In this study, we propose a reduced-order modeling framework that combines Proper Orthogonal Decomposition (POD) with a Transformer-based deep learning model to reconstruct and forecast key hemodynamic indicators—velocity components, wall shear stress (WSS), oscillatory shear index (OSI), and pressure (P)—across the pulsatile cardiac cycle. Patient-specific flow simulations were performed under age-dependent inlet conditions for two cohorts (ages 20–39 and 50–59 years) to evaluate the model's performance under physiologically realistic scenarios. POD was used to extract dominant flow features and reduce the dimensionality of the data, with energy analysis showing that fewer than 10 modes captured over 95% of the energy for velocity and pressure, while OSI required more modes due to its complex oscillatory behavior. The Transformer model, trained on the temporal evolution of the POD coefficients, accurately reproduced the flow field dynamics, achieving low reconstruction errors on both training and unseen test data. Comparative analysis revealed that the older age group exhibited more stable OSI behavior, resulting in improved predictive accuracy. In contrast, the younger group, characterized by more dynamic and energetic flow, required more modes and exhibited higher prediction error—particularly in OSI. Overall, the proposed POD + Transformer framework enables fast and reliable prediction of critical hemodynamic factors with a significant reduction in computational cost. This approach holds promise for large-scale aneurysm studies and real-time clinical applications in personalized risk assessment and treatment planning.
To our knowledge, there is no flat-histogram algorithm to sample the stationary distribution of non-equilibrium stochastic processes. The present work addresses this gap by introducing a generalization of the Wang-Landau algorithm, applied to non-equilibrium stochastic processes with local transitions. The main idea is to sample macroscopic states using a kinetic Monte Carlo algorithm to generate trial moves, which are accepted or rejected with a probability that depends inversely on the stationary distribution. The stationary distribution is refined through the simulation by a modification factor. A visitation histogram is also accumulated, and the modification factor is updated when the histogram satisfies a flatness condition. An estimate of the stationary distribution is obtained in the limit where the modification factor reaches a threshold value close to unity. To test the algorithm, we compare simulation results for several stochastic processes with theoretically known behavior. In addition, results from the kinetic flat-histogram algorithm are compared with standard exact stochastic simulations. We show that the kinetic flat-histogram algorithm can be applied to phase transitions in stochastic processes with bistability, which describe a wide range of phenomena such as epidemic spreading, population growth, chemical reactions, and consensus formation. With some adaptations, the kinetic flat-histogram algorithm can also be applied to stochastic models on lattices and complex networks.
Identifying structurally critical edges is fundamental to understanding and controlling the robustness of complex networks, yet existing edge-importance measures often rely on shortest-path assumptions, local heuristics, or application-specific flow information. In this paper, we propose resistor-network edge scoring (RNES), a physics-inspired, topology-driven approach that quantifies edge criticality through an electrical-network analogy: the network is modeled as a resistor network, repeatedly excited by random zero-mean current injections, and edge scores are obtained by aggregating the induced potential-drop fluctuations governed by Kirchhoff’s laws, so that edges with consistently strong responses are identified as structural bottlenecks. We evaluate RNES via targeted edge attacks on six real-world networks spanning transportation, citation, routing and social domains, where performance is assessed by dismantling curves of the largest connected component and a robustness index summarizing connectivity degradation under targeted edge removal. Experimental results show that RNES consistently achieves faster network fragmentation and the lowest robustness index across all datasets compared with representative shortest-path-based, local, flow-based and cycle-redundancy-based benchmarks.
Quantum Machine Learning has emerged as a promising paradigm for scientific computing, particularly when combined with Physics-Informed Neural Networks (PINNs) to solve partial differential equations. In this work, we introduce a Quantum Physics-Informed Neural Network (QPINN) framework for solving the Helmholtz equation in the context of scattered wavefield modeling. To overcome the numerical instabilities associated with point-source singularities commonly encountered in classical PINNs, we adopt the Lippmann-Schwinger integral formulation. The proposed hybrid architecture couples a classical neural-network encoder with a variational quantum circuit, enabling the learning of expressive, dynamically generated feature embeddings. Numerical experiments conducted on a heterogeneous velocity model demonstrate that the proposed QPINN achieves high solution accuracy while requiring significantly fewer trainable parameters compared to analogous classical neural network architectures. Our results highlight the potential of hybrid quantum-classical approaches as efficient and scalable alternatives for wavefield simulation and for solving wave-based inverse problems.
This work presents a new spectral collocation method using shifted Jacobi polynomials to solve the nonlinear inhomogeneous time-fractional FitzHugh-Nagumo differential equation numerically. By adjusting the Jacobi parameters alpha and beta (both greater than -1), the method improves accuracy and stability. We develop new operational matrices for both integer-order and Caputo fractional derivatives based on the shifted Jacobi basis, with detailed proofs. A thorough convergence analysis in the weighted L-2-norm shows the method achieves spectral accuracy. Theoretical results also show that, with the optimal choice of alpha and beta, this approach performs better than traditional Legendre-based methods, significantly reducing errors. The method is efficiently implemented and can handle different fractional orders and parameter settings.
The ability to explore a surface is biologically important for swimming bacteria. It has been demonstrated that a bacterium can swim near a planar surface for a prolonged period of time over a long distance. The eventual departure of the bacterium away from the surface has traditionally been interpreted as a classical one-dimensional Kramers problem about the angle made between bacterial orientation direction and the planar surface. In this work, we systematically study the near-wall dynamics of smooth-swimming Escherichia coli near a planar boundary through numerical simulations, from which we can extract the distance S the bacterium traveled during the period of time when it is near the surface. Our results demonstrate a characteristic travel distance S & lowast;, as the probability P(S) decays exponentially with S. Furthermore, we demonstrate that the characteristic travel distance S & lowast; can be sensitively modulated by bacterial dimensions, i.e. the characteristic sizes of bacteria. Our results show that the diverse S & lowast; spans almost two orders of magnitude for bacteria with small variations in dimensions, where an exponential dependence is obtained.
This study investigates the effects of wall wettability on vortex shedding behavior and hydrodynamic force coefficients around a circular pier through high-fidelity simulations using OpenFOAM. By systematically varying the static Wall Contact Angle (WCA) across a range of Reynolds numbers, the impact of surface wettability on the lift (Cl), drag (Cd) and moment (Cm) coefficients, as well as wake dynamics, is analyzed. The results demonstrate that increasing hydrophobicity significantly alters the near-wake structure by delaying flow separation, reducing vortex strength and suppressing vortex-induced vibrations. Superhydrophobic surfaces, in particular, yield narrower wakes, lower pressure drag and reduced oscillatory loads. Power spectral analyses confirm a shift or suppression of the dominant vortex shedding frequency with increased WCA. These findings highlight surface wettability as a passive, scalable strategy for controlling unsteady flow phenomena and improving structural performance in marine and civil engineering applications.
This study presents a comprehensive first-principles investigation of the structural, electronic and optical and mechanical properties of the Ag-based oxide perovskite AgYbO 3 , conducted within the framework of density functional theory (DFT) using the CASTEP computational code. The compound is found to stabilize in a highly symmetric cubic phase with the [Formula: see text] space group. Structural optimization and electronic structure calculations were performed using the GGA-RPBE functional, which revealed that AgYbO 3 exhibits semiconducting behavior with a direct band gap. This property underscores its potential suitability for low-energy electronic components, where minimal energy loss is desired. The partial density of states (PDOS) analysis provides detailed insights into the contribution of Ag, Yb and O orbital to the valence and conduction bands, particularly near the Fermi level. Optical properties were also evaluated, with the absorption spectrum showing a strong response in the ultraviolet (UV) range, thereby indicating the materials’ applicability in UV detection and optoelectronic systems. As demonstrated by the positive values of the shear modulus and the tetragonal shear constant C[Formula: see text], as well as the high Pugh’s ratio, which is larger than 1.75 and suggests ductile behavior, the computed elastic parameters also support the compound’s mechanical stability. Furthermore, low elastic anisotropy is indicated by a low Zener anisotropy factor value and considerable resistance to shear strain in the principal crystallographic axes is indicated by a high [Formula: see text] value. Collectively, these results contribute to a deeper understanding of AgYbO 3 intrinsic physical characteristics and suggest its viability as a candidate material for next-generation UV-sensitive devices and energy-efficient optoelectronic applications.
This study numerically investigates the hydrodynamic behavior and energy dissipation generated by the motion of an underwater vehicle operating above rough coastal beds using a coupled Reynolds-Averaged Navier-Stokes (RANS) and Volume of Fluid (VOF) Computational Fluid Dynamics (CFDs) framework. Unlike prior studies that mainly addressed free-surface interactions or smooth seabed conditions, this work models realistic coastal roughness through sinusoidal seabed topography to analyze its influence on near-bed flow dynamics, vortex formation and wake turbulence. Simulations were performed across multiple vehicle velocities and roughness ratios to quantify variations in hydrodynamic forces, pressure fields and energy dissipation mechanisms. Results demonstrate that seabed roughness substantially amplifies turbulence intensity, broadens wake structures and increases drag and energy loss, driven by asymmetric pressure fields and intensified vortex shedding, especially in the near-bed region. The research fills a major gap in underwater vehicle CFD studies by establishing an integrated modeling framework that captures coupled vehicle-seabed-free-surface interactions, providing novel insights into the effects of seabed-induced turbulence on vehicle performance, energy efficiency and maneuverability in complex coastal environments.
We introduce a decoupled submerged Stokeslet formulation for rigid-body Stokes flow in which the number of force points n is allowed to differ from the number of boundary collocation points N. Unlike the conventional Method of Fundamental Solutions (MFS), where the number of discretization points coincide, this approach allows fewer force unknowns (n < N) and determines them in a least-squares sense. This structural decoupling acts as a discretization-level regularization, reducing condition numbers and suppressing nonphysical force oscillations without modifying the Stokeslet kernel. Systematic benchmarks on spheres, spheroids and asymmetric geometries show that moderate force reduction combined with appropriate submerging depth yields stable second-to third-order convergence. Compared with both conventional MFS and the Method of Regularized Stokeslets (MRS), the proposed formulation achieves improved conditioning and comparable or higher accuracy for a given boundary resolution, while retaining the simplicity of global Stokeslet representations.
Coronary artery bypass grafting (CABG) is the first-line therapy for 90% coronary artery stenosis, yet conventional hemodynamic simulations of CABG ignore the microcirculation system, leading to a lack of physiological condition. In this work, we construct numerical models of one-way/two-way arterial bypass with 20 degrees, 30 degrees and 45 degrees anastomotic angles for 90% stenosis, integrate microcirculation as a seepage outlet via porous media, and adopt two-way fluid-structure interaction (FSI) to assess key hemodynamic and structural parameters across representative cross-sections. The results show that the 30 degrees anastomotic angle outperforms 20 degrees and 45 degrees in both configurations, and the two-way bypass exhibits superior hemodynamic performance compared to one-way bypass, with the 30 degrees two-way bypass model achieving the most stable structural response and optimal local hemodynamic microenvironment. This study reveals the optimal CABG design parameters under microcirculatory regulation, offering a hemodynamic theoretical basis for individualized CABG surgical planning and preoperative decision-making in cardiovascular clinical practice.
We study financial systems via multiscale topological analysis based on algebraic topology. By mapping coarse-grained time series into complex networks, we find that a universal scaling law behavior emerges between the number of higher-order cliques and the temporal scale. Remarkably, we show that the associated exponents exhibit a monotonic growth pattern irrespective of the kinds of financial system. Furthermore, we show that the multiscale entropy of financial systems presents a uniform decreasing trend. Our work, for the first time, uncovers higher-order structural organization underlying financial markets from a multiscale perspective.
In the increasingly complex domain of power systems, where load demand fluctuates due to renewable energy adoption and shifting consumption patterns, accurate short-term electric power load forecasting is essential for grid stability. Traditional forecasting models often fail to address the dynamic and nonlinear nature of load data. This study introduces a novel approach utilizing a Radial Basis Function Neural Network (RBFNN) optimized by the Multi-Objective Cuckoo Search (MOCS) algorithm. The MOCS algorithm combines the Levy flight search strategy with Pareto optimization, effectively navigating the optimization landscape to enhance the RBFNN’s forecasting accuracy while maintaining computational efficiency. The experimental results demonstrate that the MOCS-optimized RBFNN model significantly outperforms traditional models, achieving lower prediction errors and faster convergence. This advancement in forecasting methodology underscores the potential of integrating multi-objective optimization techniques in neural network models for improving the accuracy and efficiency of short-term load forecasting.
Scrub typhus is a vector-borne infectious disease transmitted to humans through infected chigger mites, with animals serving as important reservoir hosts. This work develops and analyzes a deterministic compartmental model describing the transmission dynamics of scrub typhus among humans, mites and animals. The model incorporates both horizontal transmission between interacting populations and vertical transmission within the mite population. Positivity and boundedness of solutions are established to ensure well-posedness of the model. The basic reproduction number, R-0, is derived, and the disease-free equilibrium is shown to be asymptotically stable when R-0<1. Further analysis establishes the existence of forward bifurcation. To improve the epidemiological relevance of the study, model parameters are estimated by fitting the infected human population to reported scrub typhus incidence data from Kerala, India, for the period 2014-2023. Sensitivity analysis identifies the key parameters influencing disease transmission. An optimal control framework incorporating vector control, reservoir control, personal protection and treatment is then investigated. Numerical simulations show that personal protection significantly reduces human infections, while vector control effectively suppresses overall transmission. The combined implementation of all control measures is found to be the most effective strategy. The novelty of the study lies in integrating vertical transmission, data-driven parameter estimation and multiple optimal control strategies within a unified scrub typhus modeling framework, thereby providing useful insights for public health planning and disease management.
This study evaluates the hemodynamic effects of endovascular coiling on middle cerebral artery (MCA) aneurysms in two age groups (20-39 and 60-69 years), focusing on flow-related parameters such as wall shear stress (WSS), pressure, time-averaged wall shear stress (AWSS) and oscillatory shear index (OSI). Computational fluid dynamics (CFD) simulations were conducted using age-specific pulsatile flow waveforms on patient-derived MCA aneurysm models. Computational fluid dynamics (CFD) simulations were performed using the incompressible Navier-Stokes equations with a Casson nonNewtonian blood model, rigid-wall assumption and pulsatile inlet waveforms. The coil mass was modeled as a porous medium via the Kozeny-Carman permeability formulation. Two configurations were compared: nonendovascular coiling and post-endovascular coiling. Hemodynamic variables were assessed through contour plots and quantified in terms of maximum and mean values across the aneurysm wall. Endovascular coiling significantly reduced mean WSS (from similar to 6.7Pa to similar to 2.9Pa), mean AWSS (from similar to 1170Pa to similar to 440Pa) and maximum OSI (from 0.203-0.267 to 0.070-0.086) across both age groups. Contour analyses revealed substantial disruption of intra-aneurysmal vortices and oscillatory flow patterns post-coiling, especially at the dome and neck. Despite localized increases in maximum AWSS and pressure, the overall hemodynamic burden on the aneurysm wall was markedly reduced. Age-related differences were modest, with older individuals showing slightly elevated OSI and lower pressure values. Endovascular coiling provides effective hemodynamic stabilization in MCA aneurysms by suppressing deleterious flow patterns and shear-related stresses. The treatment is beneficial across age groups, supporting its clinical utility for mitigating aneurysm rupture risk. These findings may inform patient-specific treatment planning and long-term outcome assessment.
