Computer Physics Communications最新文献

{"title":"Path-integral molecular dynamics with actively-trained and universal machine learning force fields","authors":"A.A. Solovykh ,&nbsp;N.E. Rybin ,&nbsp;I.S. Novikov ,&nbsp;A.V. Shapeev","doi":"10.1016/j.cpc.2025.109902","DOIUrl":"10.1016/j.cpc.2025.109902","url":null,"abstract":"<div><div>Accounting for nuclear quantum effects (NQEs) can significantly alter material properties at finite temperatures. Atomic modeling using the path-integral molecular dynamics (PIMD) method can fully account for such effects, but requires computationally efficient and accurate models of interatomic interactions. Empirical potentials are fast but may lack sufficient accuracy, whereas quantum-mechanical calculations are highly accurate but computationally expensive. Machine-learned interatomic potentials offer a solution to this challenge, providing near-quantum-mechanical accuracy while maintaining high computational efficiency compared to density functional theory (DFT) calculations. In this context, an interface was developed to integrate moment tensor potentials (MTPs) from the MLIP-2 software package into PIMD calculations using the i-PI software package (the MTP-PIMD approach). This interface was then applied to active learning of potentials and to investigate the influence of NQEs on material properties, namely the temperature dependence of lattice parameters and linear lattice thermal expansion (LTE) coefficients, as well as radial distribution functions, for lithium hydride (LiH) and silicon (Si) systems. The dependencies of the linear LTE on temperature for LiH and Si obtained with MTPs are in a good agreement with available experimental dependencies and with the ones calculated with the MatterSim universal machine learning force field and with the quasi-harmonic approximation. The MTP-PIMD approach thus proves to be highly accurate and effective.</div></div>","PeriodicalId":285,"journal":{"name":"Computer Physics Communications","volume":"319 ","pages":"Article 109902"},"PeriodicalIF":3.4,"publicationDate":"2025-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145360922","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Heterogeneous CPU-GPU parallelization for nonlinear coupled constitutive relations in hypersonic rarefied non-equilibrium flows","authors":"Shuhua Zeng ,&nbsp;Shaobo Yao ,&nbsp;Junyuan Yang ,&nbsp;Wenwen Zhao ,&nbsp;Jiaqi An ,&nbsp;Weifang Chen","doi":"10.1016/j.cpc.2025.109905","DOIUrl":"10.1016/j.cpc.2025.109905","url":null,"abstract":"<div><div>The nonlinear coupled constitutive relations (NCCR) have been proven to be a promising tool for rarefied non-equilibrium flows. To further optimize the efficiency of the NCCR solution in hypersonic complex flows, the first migration of the NCCR method to a graphics processing unit (GPU) platform is conducted in this study, with the application of compute unified device architecture (CUDA) and message passing interface (MPI) models. Concurrently, the data parallel lower upper relaxation (DPLUR) implicit scheme is applied to avoid data dependence during the computational processes. After a code validation, three numerical cases, i.e., hypersonic flows around a blunt cylinder with varying mesh sizes, a circular cylinder with different Knudsen numbers, and a hypersonic technology vehicle (HTV)-type flying vehicle with various Mach numbers, were investigated for assessing the performance improvement of GPU-NCCR solver on a CPU-GPU heterogeneous parallel computing platform. The results in this work show that the proposed GPU-accelerated NCCR solver on a single NVIDIA GeForce RTX 4090 GPU can achieve one or two orders of magnitude speedups, ranging from 54.5 to 179.3, in comparison to the CPU-only NCCR solution on a single AMD EPYC 7T83 CPU core. Within the computing power capabilities, the GPU-NCCR algorithm's performance gain is greater with a larger mesh size, and slightly affected by the incoming flow conditions. More importantly, the GPU-accelerated NCCR solution attains more speedups than the GPU-NS solver in hypersonic non-equilibrium flows. These superior advantages of the proposed GPU-accelerated computational strategy are expected to render the NCCR method a fairly efficient engineering approach for modelling rarefied non-equilibrium flows around hypersonic vehicles.</div></div>","PeriodicalId":285,"journal":{"name":"Computer Physics Communications","volume":"318 ","pages":"Article 109905"},"PeriodicalIF":3.4,"publicationDate":"2025-10-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145359023","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Competitive algorithms for calculating the ground state properties of Bose-Fermi mixtures","authors":"Tomasz Świsłocki ,&nbsp;Krzysztof Gawryluk ,&nbsp;Mirosław Brewczyk ,&nbsp;Tomasz Karpiuk","doi":"10.1016/j.cpc.2025.109897","DOIUrl":"10.1016/j.cpc.2025.109897","url":null,"abstract":"<div><div>In this work we define, analyze, and compare different numerical schemes that can be used to study the ground state properties of Bose-Fermi systems, such as mixtures of different atomic species under external forces or self-bound quantum droplets. The bosonic atoms are assumed to be condensed and are described by the generalized Gross-Pitaevskii equation. The fermionic atoms, on the other hand, are treated individually, and each atom is associated with a wave function whose evolution follows the Hartree-Fock equation. We solve such a formulated set of equations using a variety of methods, including those based on adiabatic switching of interactions and the imaginary time propagation technique combined with the Gram-Schmidt orthonormalization or the diagonalization of the Hamiltonian matrix. We show how different algorithms compete at the numerical level by studying the mixture in the range of parameters covering the formation of self-bound quantum Bose-Fermi droplets.</div></div>","PeriodicalId":285,"journal":{"name":"Computer Physics Communications","volume":"318 ","pages":"Article 109897"},"PeriodicalIF":3.4,"publicationDate":"2025-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145333699","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"QE-CONVERSE: An open-source package for the quantum ESPRESSO distribution to compute non-perturbatively orbital magnetization from first principles, including NMR chemical shifts and EPR parameters","authors":"S. Fioccola ,&nbsp;L. Giacomazzi ,&nbsp;D. Ceresoli ,&nbsp;N. Richard ,&nbsp;A. Hemeryck ,&nbsp;L. Martin-Samos","doi":"10.1016/j.cpc.2025.109891","DOIUrl":"10.1016/j.cpc.2025.109891","url":null,"abstract":"<div><div>Orbital magnetization, a key property arising from the orbital motion of electrons, plays a crucial role in determining the magnetic behavior of molecules and solids. Despite its straightforward calculation in finite systems, the computation in periodic systems poses challenges due to the ill-defined position operator and surface current contributions. The modern theory of orbital magnetization, formulated in the Wannier representation and implemented within the Density Functional Theory (DFT) framework, offers an accurate solution through the “converse approach.” In this paper, we introduce <figure><img></figure>, a refactored and modular implementation of the converse method, designed to replace the outdated routines from Quantum ESPRESSO (version 3.2). <figure><img></figure> integrates recent advancements in computational libraries, including scaLAPACK and ELPA, to enhance scalability and computational efficiency, particularly for large supercell calculations. While <figure><img></figure> incorporates these improvements for scalability, the main focus of this work is provide the community with a performing and accurate first principles orbital magnetization package to compute properties such as Electron Paramagnetic Resonance (EPR) g-tensors and Nuclear Magnetic Resonance (NMR) chemical shifts, specially in systems where perturbative methods fail. We demonstrate the effectiveness of <figure><img></figure> through several benchmark cases, including the NMR chemical shift of ²⁷Al in alumina and ¹⁷O and ²⁹Si in <em>α</em>-quartz, as well as the EPR g-tensor of <span><math><mmultiscripts><mrow><mi>Σ</mi></mrow><mprescripts></mprescripts><none></none><mrow><mi>n</mi></mrow></mmultiscripts><mo>(</mo><mi>n</mi><mo>≥</mo><mn>2</mn><mo>)</mo></math></span> radicals and substitutional nitrogen defects in silicon. In all cases, the results show excellent agreement with theoretical and experimental data, with significant improvements in accuracy for EPR calculations over the linear response approach. The <figure><img></figure> package, fully compatible with the latest Quantum ESPRESSO versions, opens new possibilities for studying complex materials with enhanced precision.</div></div><div><h3>Program summary</h3><div><em>Program Title:</em> qe-converse</div><div><em>CPC Library link to program files:</em> <span><span>https://doi.org/10.17632/3tyhmxknfc.1</span><svg><path></path></svg></span></div><div><em>Developer's repository link:</em> <span><span>https://github.com/mammasmias/QE-CONVERSE.git</span><svg><path></path></svg></span></div><div><em>Licensing provisions:</em> GNU General Public Licence 3.0</div><div><em>Programming language:</em> Fortran 90</div><div><em>Nature of problem:</em> Ab-initio calculation of the EPR <em>g</em>-tensor and the NMR chemical shift in solid state.</div><div><em>Solution method:</em> Compute the orbital magnetization through a non-pertubative method.</div></div>","PeriodicalId":285,"journal":{"name":"Computer Physics Communications","volume":"318 ","pages":"Article 109891"},"PeriodicalIF":3.4,"publicationDate":"2025-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145263525","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"STORM: Scrape-off layer turbulence in tokamak fusion reactors","authors":"J.T. Omotani ,&nbsp;D. Dickinson ,&nbsp;B.D. Dudson ,&nbsp;L. Easy ,&nbsp;D. Hoare ,&nbsp;P. Hill ,&nbsp;T. Nicholas ,&nbsp;J. Parker ,&nbsp;F. Riva ,&nbsp;N.R. Walkden ,&nbsp;Q. Xia ,&nbsp;F. Militello","doi":"10.1016/j.cpc.2025.109893","DOIUrl":"10.1016/j.cpc.2025.109893","url":null,"abstract":"&lt;div&gt;&lt;div&gt;The scrape-off layer of a tokamak fusion reactor carries the plasma exhaust from the hot core plasma to the material surfaces of the reactor vessel. The heat loads imposed by the exhaust are a critical limit on the performance of fusion power plants. Turbulent transport of the plasma regulates the width of the scrape-off layer plasma and must be modelled to understand the intensity of these heat loads.&lt;/div&gt;&lt;div&gt;STORM is a plasma turbulence code capable of simulating three dimensional turbulence across the full scrape-off layer of a tokamak fusion reactor, using a drift reduced, collisional fluid model. STORM uses mostly finite difference schemes, with a staggered grid in the direction parallel to the magnetic field. We describe the model, geometry and initialisation options used by STORM, as well as the numerical methods, which are implemented using the BOUT++ plasma simulation framework.&lt;/div&gt;&lt;div&gt;BOUT++ has been enhanced alongside the development of STORM, providing better support for staggered grid methods. We summarise these enhancements, including a detailed explanation of the parallel derivative methods, which underwent a major update for version 4 of BOUT++.&lt;/div&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Program summary&lt;/h3&gt;&lt;div&gt;&lt;em&gt;Program Title:&lt;/em&gt; STORM&lt;/div&gt;&lt;div&gt;&lt;em&gt;CPC Library link to program files:&lt;/em&gt; &lt;span&gt;&lt;span&gt;https://doi.org/10.17632/zm3tdfhp9r.1&lt;/span&gt;&lt;svg&gt;&lt;path&gt;&lt;/path&gt;&lt;/svg&gt;&lt;/span&gt;&lt;/div&gt;&lt;div&gt;&lt;em&gt;Developer's repository link:&lt;/em&gt; &lt;span&gt;&lt;span&gt;https://github.com/boutproject/STORM&lt;/span&gt;&lt;svg&gt;&lt;path&gt;&lt;/path&gt;&lt;/svg&gt;&lt;/span&gt;&lt;/div&gt;&lt;div&gt;&lt;em&gt;Licensing provisions:&lt;/em&gt; GPLv3&lt;/div&gt;&lt;div&gt;&lt;em&gt;Programming language:&lt;/em&gt; C++&lt;/div&gt;&lt;div&gt;&lt;em&gt;Supplementary material:&lt;/em&gt; Configuration and input files and post-processing scripts to run the example code given in Listings 1, 2, and 3.&lt;/div&gt;&lt;div&gt;&lt;em&gt;Nature of problem:&lt;/em&gt; The scrape-off layer region of tokamak fusion reactors carries the plasma exhaust which escapes from the core, confined plasma and reaches material surfaces along open magnetic field lines. The power and particle loads on the material surfaces are a critical limiting factor for the performance of fusion reactors, but are challenging to simulate due to the large fluctuation amplitudes, complex magnetic geometry, and widely separated time- and length-scales. Three dimensional simulations of plasma turbulence are needed to understand the particle and energy transport in the scrape-off layer and provide predictive capability for the design of future reactors.&lt;/div&gt;&lt;div&gt;&lt;em&gt;Solution method:&lt;/em&gt; STORM solves a drift reduced, collisional, fluid model for the scrape-off layer plasma. The model is discretised in space using mostly finite difference methods, combined in some places with Fourier methods that take advantage of the toroidal symmetry of the tokamak geometry. The fastest dynamics occur in the direction parallel to the magnetic field, for which a staggered grid is used to avoid the chequerboard instability associated with advective equations [1, s","PeriodicalId":285,"journal":{"name":"Computer Physics Communications","volume":"318 ","pages":"Article 109893"},"PeriodicalIF":3.4,"publicationDate":"2025-10-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145263437","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Comparative analysis of Richardson-Lucy deconvolution and data unfolding with mean integrated square error optimization","authors":"Nikolay D. Gagunashvili","doi":"10.1016/j.cpc.2025.109894","DOIUrl":"10.1016/j.cpc.2025.109894","url":null,"abstract":"<div><div>Two maximum likelihood-based algorithms for unfolding or deconvolution are considered: the Richardson-Lucy method and the Data Unfolding method with Mean Integrated Square Error (MISE) optimization. Unfolding is viewed as a procedure for estimating an unknown probability density function. Both external and internal quality assessment methods can be applied for this purpose. In some cases, external criteria exist to evaluate deconvolution quality. A typical example is the deconvolution of a blurred image, where the sharpness of the restored image serves as an indicator of quality. However, defining such external criteria can be challenging, particularly when a measurement has not been performed previously. In such instances, internal criteria are necessary to assess the quality of the result independently of external information. The article discusses two internal criteria: MISE for the unfolded distribution and the condition number of the correlation matrix of the unfolded distribution. These internal quality criteria are applied to a comparative analysis of the two methods using identical numerical data. The results of the analysis demonstrate the superiority of the Data Unfolding method with MISE optimization over the Richardson-Lucy method.</div></div>","PeriodicalId":285,"journal":{"name":"Computer Physics Communications","volume":"318 ","pages":"Article 109894"},"PeriodicalIF":3.4,"publicationDate":"2025-10-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145263441","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Hierarchical autoregressive neural networks in three-dimensional statistical system","authors":"Piotr Białas ,&nbsp;Vaibhav Chahar ,&nbsp;Piotr Korcyl ,&nbsp;Tomasz Stebel ,&nbsp;Mateusz Winiarski ,&nbsp;Dawid Zapolski","doi":"10.1016/j.cpc.2025.109892","DOIUrl":"10.1016/j.cpc.2025.109892","url":null,"abstract":"<div><div>Autoregressive Neural Networks (ANN) have been recently proposed as a mechanism to improve the efficiency of Monte Carlo algorithms for several spin systems. The idea relies on the fact that the total probability of a configuration can be factorized into conditional probabilities of each spin, which in turn can be approximated by a neural network. Once trained, the ANNs can be used to sample configurations from the approximated probability distribution and to explicitly evaluate this probability for a given configuration. It has also been observed that such conditional probabilities give access to information-theoretic observables such as mutual information or entanglement entropy. In this paper, we describe the hierarchical autoregressive network (HAN) algorithm in three spatial dimensions and study its performance using the example of the Ising model. We compare HAN with three other autoregressive architectures and the classical Wolff cluster algorithm. Finally, we provide estimates of thermodynamic observables for the three-dimensional Ising model, such as entropy and free energy, in a range of temperatures across the phase transition.</div></div>","PeriodicalId":285,"journal":{"name":"Computer Physics Communications","volume":"318 ","pages":"Article 109892"},"PeriodicalIF":3.4,"publicationDate":"2025-10-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145263442","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Converting sWeights to probabilities with density ratios","authors":"D.I. Glazier ,&nbsp;R. Tyson","doi":"10.1016/j.cpc.2025.109890","DOIUrl":"10.1016/j.cpc.2025.109890","url":null,"abstract":"<div><div>The use of machine learning approaches continues to have many benefits in experimental nuclear and particle physics. One common issue is generating training data which is sufficiently realistic to give reliable results. Here we advocate using real experimental data as the source of training data and demonstrate how one might subtract background contributions through the use of probabilistic weights which can be readily applied to training data. The <em>sPlot</em> formalism is a common tool used to isolate distributions from different sources. However, the negative <em>sWeights</em> produced by the <em>sPlot</em> technique can cause training problems and poor predictive power. This article demonstrates how density ratio estimation can be applied to convert <em>sWeights</em> to event probabilities, which we call <em>drWeights</em>. The <em>drWeights</em> can then be applied to produce the distributions of interest and are consistent with direct use of the <em>sWeights</em>. This article will also show how decision trees are particularly well suited to convert <em>sWeights</em>, with the benefit of fast prediction rates and adaptability to aspects of experimental data such as the data sample size and proportions of different event sources. We also show that a density ratio product approach in which the initial <em>drWeights</em> are reweighted by an additional converter gives substantially better results.</div></div>","PeriodicalId":285,"journal":{"name":"Computer Physics Communications","volume":"318 ","pages":"Article 109890"},"PeriodicalIF":3.4,"publicationDate":"2025-10-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145263440","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Fire: An open-source adaptive mesh refinement solver for supersonic reacting flows","authors":"E. Fan ,&nbsp;Tianhan Zhang ,&nbsp;Jiaao Hao ,&nbsp;Chih-Yung Wen ,&nbsp;Lisong Shi","doi":"10.1016/j.cpc.2025.109881","DOIUrl":"10.1016/j.cpc.2025.109881","url":null,"abstract":"<div><div>In this study, we introduce Fire, an open-source adaptive mesh refinement (AMR) solver for supersonic reacting flows, and conduct theoretical analyses on the efficiency of AMR methods. Fire is developed within the AMR framework of ECOGEN (Schmidmayer et al., 2020). To accurately model compressible multi-component reacting flows, the Fire solver employs the thermally perfect gas model for multi-species gaseous mixtures, mixture-averaged transport models for viscous fluxes, and detailed finite-rate chemistry for combustion processes. The solver utilizes the Harten-Lax-van Leer Contact approximate Riemann solver with low-Mach number correction to evaluate inviscid fluxes, demonstrating its superiority over the traditional Harten-Lax-van Leer Contact solver on detonation simulations. Moreover, we deduce the theoretical speedup ratio (denoted as <span><math><msub><mrow><mi>η</mi></mrow><mrow><mtext>the</mtext></mrow></msub></math></span>) of AMR methods over uniform-grid methods by analyzing the advancing procedures. This theoretical analysis is well-supported by the numerical speedup ratio (denoted as <span><math><msub><mrow><mi>η</mi></mrow><mrow><mtext>num</mtext></mrow></msub></math></span>) given by numerical tests. To further enhance computational efficiency, we propose a three-stage AMR strategy specifically tailored to the characteristics of inert flows, flame fronts, and shock-flame interactions. Comprehensive validation tests, encompassing unsteady convection and diffusion, planar deflagration, inert and reacting shock-bubble interactions, planar detonations, and detonation cellular structures, confirm the accuracy and efficiency of Fire in simulating supersonic combustions. We anticipate that this work will not only serve as a valuable numerical tool for supersonic reacting flows research but also contribute to a deeper understanding and improvement of AMR methodologies.</div></div>","PeriodicalId":285,"journal":{"name":"Computer Physics Communications","volume":"318 ","pages":"Article 109881"},"PeriodicalIF":3.4,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145263438","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Physics informed neural networks with variable Eddington factor iteration for linear radiative transfer equations","authors":"Yuhang Wu ,&nbsp;Jianhua Huang ,&nbsp;Xu Qian ,&nbsp;Wenjun Sun","doi":"10.1016/j.cpc.2025.109879","DOIUrl":"10.1016/j.cpc.2025.109879","url":null,"abstract":"<div><div>In this paper, a Physics Informed Neural Networks (PINNs) method based on Variable Eddington Factor (VEF) acceleration iteration is proposed to address the time-dependent linear radiative transfer equations (LRTEs), which exhibit the characteristics of multi-scale and high dimensionality. Firstly, the factors relating to the failure of the vanilla PINNs in solving LRTEs within the diffusion regime are analyzed by the Neural Tangent Kernel (NTK) theory. Subsequently, the VEF-PINNs method is established, where PINNs are employed to handle the radiative transfer equations and the analytic VEF equations that are used to accelerate the iteration process. It is demonstrated that as the Knudsen number <em>ε</em> approaches 0, the VEF-PINNs method converges to the iteration of diffusion limit equations, thereby ensuring the proposed method maintains the asymptotic preserving property. A theoretical analysis about the approximation errors of the iterative solution of the VEF-PINNs method is given. To evaluate the performance of the method, comparisons are made with the vanilla PINNs and Asymptotic Preserving Neural Networks (APNNs) based on micro-macro decomposition. The results reveal that the proposed VEF-PINNs can effectively solve LRTEs in various opacity regimes and can enhance the solving efficiency to a certain extent.</div></div>","PeriodicalId":285,"journal":{"name":"Computer Physics Communications","volume":"318 ","pages":"Article 109879"},"PeriodicalIF":3.4,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145263524","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}