{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,14]],"date-time":"2026-07-14T16:01:58Z","timestamp":1784044918180,"version":"3.55.0"},"reference-count":50,"publisher":"MDPI AG","issue":"9","license":[{"start":{"date-parts":[[2024,8,29]],"date-time":"2024-08-29T00:00:00Z","timestamp":1724889600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"National Natural Science Foundation of China","award":["42376194"],"award-info":[{"award-number":["42376194"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Axioms"],"abstract":"<jats:p>The Boltzmann equation with multiple-relaxation-time (MRT) collision operators has been widely employed in kinetic theory to describe the behavior of gases and liquids at the macro-level. Given the successful development of deep learning and the availability of data analytic tools, it is a feasible idea to try to solve the Boltzmann-MRT equation using a neural network-based method. Based on the canonical polyadic decomposition, a new physics-informed neural network describing the Boltzmann-MRT equation, named the network for MRT collision (NMRT), is proposed in this paper for solving the Boltzmann-MRT equation. The method of tensor decomposition in the Boltzmann-MRT equation is utilized to combine the collision matrix with discrete distribution functions within the moment space. Multiscale modeling is adopted to accelerate the convergence of high frequencies for the equations. The micro\u2013macro decomposition method is applied to improve learning efficiency. The problem-dependent loss function is proposed to balance the weight of the function for different conditions at different velocities. These strategies will greatly improve the accuracy of the network. The numerical experiments are tested, including the advection\u2013diffusion problem and the wave propagation problem. The results of the numerical simulation show that the network-based method can obtain a measure of accuracy at O10\u22123.<\/jats:p>","DOI":"10.3390\/axioms13090588","type":"journal-article","created":{"date-parts":[[2024,8,29]],"date-time":"2024-08-29T08:01:47Z","timestamp":1724918507000},"page":"588","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":5,"title":["A Physics-Informed Neural Network Based on the Boltzmann Equation with Multiple-Relaxation-Time Collision Operators"],"prefix":"10.3390","volume":"13","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-6554-9617","authenticated-orcid":false,"given":"Zhixiang","family":"Liu","sequence":"first","affiliation":[{"name":"College of Information Technology, Shanghai Ocean University, Shanghai 201306, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0004-1583-0506","authenticated-orcid":false,"given":"Chenkai","family":"Zhang","sequence":"additional","affiliation":[{"name":"College of Information Technology, Shanghai Ocean University, Shanghai 201306, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9656-9781","authenticated-orcid":false,"given":"Wenhao","family":"Zhu","sequence":"additional","affiliation":[{"name":"School of Computer Engineering and Science, Shanghai University, Shanghai 200444, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5698-8279","authenticated-orcid":false,"given":"Dongmei","family":"Huang","sequence":"additional","affiliation":[{"name":"College of Electronic and Information Engineering, Shanghai University of Electric Power, Shanghai 201306, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2024,8,29]]},"reference":[{"key":"ref_1","first-page":"128","article-title":"Deep learning vs. traditional computer vision","volume":"Volume 943","author":"Arai","year":"2020","journal-title":"Advances in Computer Vision, Proceedings of the 2019 Computer Vision Conference (CVC), Las Vegas, NV, USA, 25\u201326 April 2019"},{"key":"ref_2","unstructured":"Chowdhary, K., and Chowdhary, K. (2022). Natural language processing. Fundamentals of Artificial Intelligence, Springer."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"30033","DOI":"10.1073\/pnas.1907373117","article-title":"The unreasonable effectiveness of deep learning in artificial intelligence","volume":"117","author":"Sejnowski","year":"2020","journal-title":"Proc. Natl. Acad. Sci. USA"},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"176","DOI":"10.1038\/s41386-020-0767-z","article-title":"Deep learning for small and big data in psychiatry","volume":"46","author":"Koppe","year":"2021","journal-title":"Neuropsychopharmacology"},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"209","DOI":"10.1016\/S0376-0421(02)00005-2","article-title":"Verification and validation in computational fluid dynamics","volume":"38","author":"Oberkampf","year":"2002","journal-title":"Prog. Aerosp. Sci."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"11","DOI":"10.1016\/j.actaastro.2020.05.021","article-title":"Supervised learning mixing characteristics of film cooling in a rocket combustor using convolutional neural networks","volume":"175","author":"Ma","year":"2020","journal-title":"Acta Astronaut."},{"key":"ref_7","unstructured":"Raissi, M., Perdikaris, P., and Karniadakis, G.E. (2017). Physics informed deep learning (Part I): Data-driven solutions of nonlinear partial differential equations. arXiv."},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"686","DOI":"10.1016\/j.jcp.2018.10.045","article-title":"Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations","volume":"378","author":"Raissi","year":"2019","journal-title":"J. Comput. Phys."},{"key":"ref_9","doi-asserted-by":"crossref","unstructured":"Guo, Y., Cao, X., Liu, B., and Gao, M. (2020). Solving partial differential equations using deep learning and physical constraints. Appl. Sci., 10.","DOI":"10.3390\/app10175917"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"113250","DOI":"10.1016\/j.cma.2020.113250","article-title":"PPINN: Parareal physics-informed neural network for time-dependent PDEs","volume":"370","author":"Meng","year":"2020","journal-title":"Comput. Methods Appl. Mech. Eng."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"124700","DOI":"10.1016\/j.jhydrol.2020.124700","article-title":"Deep learning of subsurface flow via theory-guided neural network","volume":"584","author":"Wang","year":"2020","journal-title":"J. Hydrol."},{"key":"ref_12","first-page":"23761","article-title":"Separable physics-informed neural networks","volume":"36","author":"Cho","year":"2024","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"219","DOI":"10.1016\/0167-2789(91)90292-H","article-title":"The lattice Boltzmann equation: A new tool for computational fluid-dynamics","volume":"47","author":"Succi","year":"1991","journal-title":"Phys. Nonlinear Phenom."},{"key":"ref_14","doi-asserted-by":"crossref","unstructured":"Li, J. (2020). Multiscale and Multiphysics Flow Simulations of Using the Boltzmann Equation, Springer. Applications to Porous Media and MEMS.","DOI":"10.1007\/978-3-030-26466-6"},{"key":"ref_15","unstructured":"Simonis, S. (2023). Lattice Boltzmann Methods for Partial Differential Equations. [Ph.D. Thesis, Karlsruher Institut f\u00fcr Technologie (KIT)]."},{"key":"ref_16","doi-asserted-by":"crossref","first-page":"026101","DOI":"10.1063\/1.2837174","article-title":"A generalized Bhatnagar\u2013Gross\u2013Krook model for nonequilibrium flows","volume":"20","author":"Xu","year":"2008","journal-title":"Phys. Fluids"},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"037134","DOI":"10.1063\/5.0046866","article-title":"A multiple-relaxation-time collision model for nonequilibrium flows","volume":"33","author":"Shi","year":"2021","journal-title":"Phys. Fluids"},{"key":"ref_18","doi-asserted-by":"crossref","unstructured":"Bird, G.A. (1994). Molecular Gas Dynamics and the Direct Simulation of Gas Flows, Oxford University Press.","DOI":"10.1093\/oso\/9780198561958.001.0001"},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"21","DOI":"10.1186\/s42774-020-00045-8","article-title":"A unified gas-kinetic scheme for micro flow simulation based on linearized kinetic equation","volume":"2","author":"Liu","year":"2020","journal-title":"Adv. Aerodyn."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"B658","DOI":"10.1137\/16M1096001","article-title":"A fast spectral method for the Boltzmann collision operator with general collision kernels","volume":"39","author":"Gamba","year":"2017","journal-title":"SIAM J. Sci. Comput."},{"key":"ref_21","doi-asserted-by":"crossref","unstructured":"Liu, Z., Chen, Y., Xiao, W., Song, W., and Li, Y. (2023). Large-Scale Cluster Parallel Strategy for Regularized Lattice Boltzmann Method with Sub-Grid Scale Model in Large Eddy Simulation. Appl. Sci., 13.","DOI":"10.3390\/app131911078"},{"key":"ref_22","first-page":"3913968","article-title":"The Immersed Boundary-Lattice boltzmann method parallel model for fluid-structure interaction on heterogeneous platforms","volume":"2020","author":"Liu","year":"2020","journal-title":"Math. Probl. Eng."},{"key":"ref_23","doi-asserted-by":"crossref","unstructured":"Liu, Z., Li, S., Ruan, J., Zhang, W., Zhou, L., Huang, D., and Xu, J. (2023). A New Multi-Level Grid Multiple-Relaxation-Time Lattice Boltzmann Method with Spatial Interpolation. Mathematics, 11.","DOI":"10.3390\/math11051089"},{"key":"ref_24","unstructured":"Li, Z., Wang, Y., Liu, H., Wang, Z., and Dong, B. (2023). Solving Boltzmann equation with neural sparse representation. arXiv."},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Oh, J., Cho, S.Y., Yun, S.B., Park, E., and Hong, Y. (2024). Separable Physics-informed Neural Networks for Solving the BGK Model of the Boltzmann Equation. arXiv.","DOI":"10.2139\/ssrn.4858127"},{"key":"ref_26","doi-asserted-by":"crossref","first-page":"110676","DOI":"10.1016\/j.jcp.2021.110676","article-title":"Physics-informed neural networks for solving forward and inverse flow problems via the Boltzmann-BGK formulation","volume":"447","author":"Lou","year":"2021","journal-title":"J. Comput. Phys."},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"015312","DOI":"10.1103\/PhysRevE.104.015312","article-title":"Multiple-relaxation-time lattice Boltzmann model-based four-level finite-difference scheme for one-dimensional diffusion equations","volume":"104","author":"Lin","year":"2021","journal-title":"Phys. Rev. E"},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"19","DOI":"10.1080\/14786446008642818","article-title":"Illustrations of the dynamical theory of gases","volume":"19","author":"Maxwell","year":"1867","journal-title":"Philos. Mag."},{"key":"ref_29","doi-asserted-by":"crossref","first-page":"115","DOI":"10.1016\/j.matcom.2023.05.019","article-title":"One-dimensional lattice Boltzmann simulation of parallel plate dielectric barrier discharge plasma in atmospheric argon","volume":"213","author":"Choe","year":"2023","journal-title":"Math. Comput. Simul."},{"key":"ref_30","doi-asserted-by":"crossref","first-page":"023310","DOI":"10.1103\/PhysRevE.94.023310","article-title":"Multiple-relaxation-time color-gradient lattice Boltzmann model for simulating two-phase flows with high density ratio","volume":"94","author":"Ba","year":"2016","journal-title":"Phys. Rev. E"},{"key":"ref_31","doi-asserted-by":"crossref","unstructured":"Huang, X., Liu, H., Shi, B., Wang, Z., Yang, K., Li, Y., Weng, B., Wang, M., Chu, H., and Zhou, J. (2021). Solving partial differential equations with point source based on physics-informed neural networks. arXiv.","DOI":"10.24963\/ijcai.2022\/533"},{"key":"ref_32","doi-asserted-by":"crossref","first-page":"51","DOI":"10.1016\/j.cjph.2023.04.006","article-title":"An improved MRT-LBM and investigation to the transition and periodicity of 2D lid-driven cavity flow with high Reynolds numbers","volume":"84","author":"Zhang","year":"2023","journal-title":"Chin. J. Phys."},{"key":"ref_33","first-page":"7462","article-title":"Implicit neural representations with periodic activation functions","volume":"33","author":"Sitzmann","year":"2020","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_34","unstructured":"Xu, Z.Q.J., Zhang, Y., Luo, T., Xiao, Y., and Ma, Z. (2019). Frequency principle: Fourier analysis sheds light on deep neural networks. arXiv."},{"key":"ref_35","doi-asserted-by":"crossref","first-page":"115","DOI":"10.1038\/s41746-019-0193-y","article-title":"Integrating machine learning and multiscale modeling\u2014perspectives, challenges, and opportunities in the biological, biomedical, and behavioral sciences","volume":"2","author":"Alber","year":"2019","journal-title":"NPJ Digit. Med."},{"key":"ref_36","doi-asserted-by":"crossref","unstructured":"Liu, Z., Cai, W., and Xu, Z.Q.J. (2020). Multi-scale deep neural network (MscaleDNN) for solving Poisson-Boltzmann equation in complex domains. arXiv.","DOI":"10.4208\/cicp.OA-2020-0179"},{"key":"ref_37","unstructured":"Edwards, T. (2024, June 11). Discrete Wavelet Transforms: Theory and Implementation. Available online: https:\/\/citeseerx.ist.psu.edu\/document?repid=rep1&type=pdf&doi=f7efbe4055f84612ec0851f6ccd11d2d4999141b."},{"key":"ref_38","doi-asserted-by":"crossref","first-page":"4580","DOI":"10.1137\/090756077","article-title":"A micro-macro decomposition-based asymptotic-preserving scheme for the multispecies Boltzmann equation","volume":"31","author":"Jin","year":"2010","journal-title":"SIAM J. Sci. Comput."},{"key":"ref_39","doi-asserted-by":"crossref","first-page":"264","DOI":"10.1016\/j.jcp.2019.01.018","article-title":"Micro-macro decomposition based asymptotic-preserving numerical schemes and numerical moments conservation for collisional nonlinear kinetic equations","volume":"382","author":"Gamba","year":"2019","journal-title":"J. Comput. Phys."},{"key":"ref_40","doi-asserted-by":"crossref","first-page":"115384","DOI":"10.1109\/ACCESS.2022.3218802","article-title":"Model-based deep learning: On the intersection of deep learning and optimization","volume":"10","author":"Shlezinger","year":"2022","journal-title":"IEEE Access"},{"key":"ref_41","doi-asserted-by":"crossref","first-page":"519","DOI":"10.1016\/j.jcp.2018.08.057","article-title":"Parallel tensor methods for high-dimensional linear PDEs","volume":"375","author":"Boelens","year":"2018","journal-title":"J. Comput. Phys."},{"key":"ref_42","doi-asserted-by":"crossref","first-page":"A2634","DOI":"10.1137\/15M1042802","article-title":"Randomized alternating least squares for canonical tensor decompositions: Application to a PDE with random data","volume":"38","author":"Reynolds","year":"2016","journal-title":"SIAM J. Sci. Comput."},{"key":"ref_43","doi-asserted-by":"crossref","unstructured":"Lange, K. (2010). Singular value decomposition. Numerical Analysis for Statisticians, Springer. Statistics and Computing.","DOI":"10.1007\/978-1-4419-5945-4"},{"key":"ref_44","doi-asserted-by":"crossref","first-page":"937","DOI":"10.1109\/LSP.2022.3156870","article-title":"Canonical Polyadic Decomposition via the generalized Schur decomposition","volume":"29","author":"Evert","year":"2022","journal-title":"IEEE Signal Process. Lett."},{"key":"ref_45","unstructured":"Peng, W., Zhou, W., Zhang, J., and Yao, W. (2020). Accelerating physics-informed neural network training with prior dictionaries. arXiv."},{"key":"ref_46","unstructured":"Loshchilov, I., and Hutter, F. (2016). Sgdr: Stochastic gradient descent with warm restarts. arXiv."},{"key":"ref_47","doi-asserted-by":"crossref","first-page":"185","DOI":"10.1016\/j.aej.2021.04.066","article-title":"A study of the generalized nonlinear advection-diffusion equation arising in engineering sciences","volume":"61","author":"Adeyemo","year":"2022","journal-title":"Alex. Eng. J."},{"key":"ref_48","doi-asserted-by":"crossref","unstructured":"Maragkos, G., and Beji, T. (2021). Review of convective heat transfer modelling in cfd simulations of fire-driven flows. Appl. Sci., 11.","DOI":"10.3390\/app11115240"},{"key":"ref_49","doi-asserted-by":"crossref","first-page":"53","DOI":"10.1017\/jfm.2014.79","article-title":"Solving the Boltzmann equation deterministically by the fast spectral method: Application to gas microflows","volume":"746","author":"Wu","year":"2014","journal-title":"J. Fluid Mech."},{"key":"ref_50","unstructured":"Sch\u00e4fer, V. (2022). Generalization of Physics-Informed Neural Networks for Various Boundary and Initial Conditions. [Ph.D. Thesis, Technische Universit\u00e4t Kaiserslautern]."}],"container-title":["Axioms"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2075-1680\/13\/9\/588\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,10]],"date-time":"2025-10-10T15:44:56Z","timestamp":1760111096000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2075-1680\/13\/9\/588"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,8,29]]},"references-count":50,"journal-issue":{"issue":"9","published-online":{"date-parts":[[2024,9]]}},"alternative-id":["axioms13090588"],"URL":"https:\/\/doi.org\/10.3390\/axioms13090588","relation":{},"ISSN":["2075-1680"],"issn-type":[{"value":"2075-1680","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,8,29]]}}}