Skip to main content
Log in

Solving multiscale steady radiative transfer equation using neural networks with uniform stability

  • Research
  • Published:
Research in the Mathematical Sciences Aims and scope Submit manuscript


This paper concerns solving the steady radiative transfer equation with diffusive scaling, using the physics informed neural networks (PINNs). The idea of PINNs is to minimize a least-square loss function, that consists of the residual from the governing equation, the mismatch from the boundary conditions, and other physical constraints such as conservation. It is advantageous of being flexible and easy to execute, and brings the potential for high dimensional problems. Nevertheless, due the presence of small scales, the vanilla PINNs can be extremely unstable for solving multiscale steady transfer equations. In this paper, we propose a new formulation of the loss based on the macro-micro decomposition. We prove that, the new loss function is uniformly stable with respect to the small Knudsen number in the sense that the \(L^2\)-error of the neural network solution is uniformly controlled by the loss. When the boundary condition is an-isotropic, a boundary layer emerges in the diffusion limit and therefore brings an additional difficulty in training the neural network. To resolve this issue, we include a boundary layer corrector that carries over the sharp transition part of the solution and leaves the rest easy to be approximated. The effectiveness of the new methodology is demonstrated in extensive numerical examples.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12
Fig. 13
Fig. 14
Fig. 15
Fig. 16
Fig. 17
Fig. 18

Similar content being viewed by others

Data availability statement

Data sharing not applicable to this article as no datasets were generated or analysed during the current study.


  1. Bardos, C., Santos, R., Sentis, R.: Diffusion approximation and computation of the critical size. Trans. Am. Math. Soc. 284, 617–649 (1984)

    Article  MathSciNet  Google Scholar 

  2. Bensoussan, A., Lions, P.-L., Papanicolaou, G.C.: Boundary layers and homogenizatlon of transport processes. Publ. Res. Inst. Math. Sci. 15, 53–157 (1979)

    Article  MathSciNet  Google Scholar 

  3. Boscarino, S., Pareschi, L., Russo, G.: Implicit-explicit Runge–Kutta scheme for hyperbolic systems and kinetic equations in the diffusion limit. SIAM J. Sci. Comput. 35, 22–51 (2013)

    Article  MathSciNet  Google Scholar 

  4. Chen, Z., Liu, L., Mu, L.: Solving the linear transport equation by a deep neural network approach, arXiv preprint arXiv:2102.09157 (2021)

  5. Dimarco, G., Pareschi, L.: Numerical methods for kinetic equations. Acta Numer. 23, 369–520 (2014)

    Article  MathSciNet  Google Scholar 

  6. Egger, H., Schlottbom, M.: An lp theory for stationary radiative transfer. Appl. Anal. 93, 1283–1296 (2014)

    Article  MathSciNet  Google Scholar 

  7. Golse, F., Jin, S., Levermore, C.D.: A domain decomposition analysis for a two-scale linear transport problem. ESAIM: Math. Model. Numer. Anal. 37, 869–892 (2003)

    Article  MathSciNet  Google Scholar 

  8. Han, H., Tang, M., Ying, W.: Two uniform tailored finite point schemes for the two dimensional discrete ordinates transport equations with boundary and interface layers. Commun. Comput. Phys. 15, 797–826 (2014)

    Article  MathSciNet  Google Scholar 

  9. Han, J., Jentzen, A., et al.: Algorithms for solving high dimensional pdes: from nonlinear monte carlo to machine learning. arXiv preprint arXiv:2008.13333 (2020)

  10. Han, J., Jentzen, A., Weinan, E.: Solving high-dimensional partial differential equations using deep learning. Proc. Natl. Acad. Sci. 115, 8505–8510 (2018)

    Article  MathSciNet  Google Scholar 

  11. Hwang, H.J., Jang, J.W., Jo, H., Lee, J.Y.: Trend to equilibrium for the kinetic Fokker-Planck equation via the neural network approach. J. Comput. Phys. 419, 109665 (2020)

    Article  MathSciNet  Google Scholar 

  12. Jin, S.: Asymptotic preserving (AP) schemes for multiscale kinetic and hyperbolic equations: a review. Riv. Mat. Univ. Parma 2, 177–216 (2012)

    MathSciNet  MATH  Google Scholar 

  13. Jin, S., Ma, Z., Wu, K.: Asymptotic-preserving neural networks for multiscale time-dependent linear transport equations, arXiv preprint arXiv:2111.02541 (2021)

  14. Jin, S., Pareschi, L., Toscani, G.: Uniformly accurate diffusive relaxation schemes for multiscale transport equations. SIAM J. Numer. Anal. 38, 913–936 (2000)

    Article  MathSciNet  Google Scholar 

  15. Jin, S., Tang, M., Han, H.: A uniformly second order numerical method for the one-dimensional discrete-ordinate transport equation and its diffusion limit with interface. Netw. Heterogeneous Media 4, 35–65 (2009)

    Article  MathSciNet  Google Scholar 

  16. Klar, A.: Asymptotic-induced domain decomposition methods for kinetic and drift diffusion semiconductor equations. SIAM J. Sci. Comput. 19, 2032–2050 (1998)

    Article  MathSciNet  Google Scholar 

  17. Klar, A.: An asymptotic-induced scheme for nonstationary transport equations in the diffusive limit. SIAM J. Numer. Anal. 35, 1073–1094 (1998)

    Article  MathSciNet  Google Scholar 

  18. Lagaris, I.E., Likas, A., Fotiadis, D.I.: Artificial neural networks for solving ordinary and partial differential equations. IEEE Trans. Neural Netw. 9, 987–1000 (1998)

    Article  Google Scholar 

  19. Lee, J. Y., Jang, J. W., Hwang, H. J.: The model reduction of the Vlasov-Poisson-Fokker-Planck system to the Poisson-Nernst-Planck system via the deep neural network approach, arXiv preprint arXiv:2009.13280 (2020)

  20. Lemou, M., Méhats, F.: Micro-macro schemes for kinetic equations including boundary layers. SIAM J. Sci. Comput. 34, B734–B760 (2012)

    Article  MathSciNet  Google Scholar 

  21. Lemou, M., Mieussens, L.: New asymptotic preserving scheme based on micro-macro formulation for linear kinetic equations in the diffusion limit. SIAM J. Sci. Comput. 31, 334–368 (2008)

    Article  MathSciNet  Google Scholar 

  22. Lewis, E., Miller, W., Jr.: Computational Methods of Neutron Transport. Wiley, London (1983)

  23. Li, Q., Lu, J., Sun, W.: Diffusion approximations and domain decomposition method of linear transport equations: asymptotics and numerics. J. Comput. Phys. 292, 141–167 (2015)

    Article  MathSciNet  Google Scholar 

  24. Li, Q., Wang, L.: Implicit asymptotic preserving method for linear transport equations. Commun. Comput. Phys. 22, 157–181 (2017)

    Article  MathSciNet  Google Scholar 

  25. Li, W., Song, P., Wang, Y.: An asymptotic-preserving imex method for nonlinear radiative transfer equation, arXiv preprint arXiv:2008.06730 (2020)

  26. Liu, L., Zeng, T., Zhang, Z.: A deep neural network approach on solving the linear transport model under diffusive scaling, arXiv preprint arXiv:2102.12408 (2021)

  27. Lu, J., Lu, Y.: A priori generalization error analysis of two-layer neural networks for solving high dimensional schrödinger eigenvalue problems, arXiv preprint arXiv:2105.01228 (2021)

  28. Lu, Y., Lu, J., Wang, M.: A priori generalization analysis of the deep ritz method for solving high dimensional elliptic partial differential equations. In: Conference on Learning Theory, PMLR, pp. 3196–3241 (2021)

  29. Manteuffel, T.A., Ressel, K.J., Starke, G.: A boundary functional for the least-squares finite-element solution of neutron transport problems. SIAM J. Numer. Anal. 37, 556–586 (1999)

    Article  MathSciNet  Google Scholar 

  30. Mishra, S., Molinaro, R.: Estimates on the generalization error of physics informed neural networks (PINNs) for approximating PDEs (2020). arXiv preprint arXiv:2006.16144

  31. Peng, Z., Cheng, Y., Qiu, J.-M., Li, F.: Stability-enhanced ap imex-ldg schemes for linear kinetic transport equations under a diffusive scaling. J. Comput. Phys. 415, 109485 (2020)

    Article  MathSciNet  Google Scholar 

  32. Raissi, M., Perdikaris, P., Karniadakis, G.E.: Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. J. Comput. Phys. 378, 686–707 (2019)

    Article  MathSciNet  Google Scholar 

  33. Shalev-Shwartz, S., Ben-David, S.: Understanding Machine Learning: From Theory to Algorithms. Cambridge University Press, Cambridge (2014)

    Book  Google Scholar 

  34. Sirignano, J., Spiliopoulos, K.: Dgm: a deep learning algorithm for solving partial differential equations. J. Comput. Phys. 375, 1339–1364 (2018)

    Article  MathSciNet  Google Scholar 

  35. Sun, W., Jiang, S., Xu, K.: An implicit unified gas kinetic scheme for radiative transfer with equilibrium and non-equilibrium diffusive limits. Commun. Comput. Phys. 22, 889–912 (2017)

    Article  MathSciNet  Google Scholar 

  36. Tang, M., Wang, L., Zhang, X.: Accurate front capturing asymptotic preserving scheme for nonlinear gray radiative transfer equation. SIAM J. Sci. Comput. 43, B759–B783 (2021)

    Article  MathSciNet  Google Scholar 

  37. Wang, S., Teng, Y., Perdikaris, P.: Understanding and mitigating gradient pathologies in physics-informed neural networks, arXiv preprint arXiv:2001.04536 (2020)

  38. Wang, S., Yu, X., Perdikaris, P.: When and why pinns fail to train: a neural tangent kernel perspective, arXiv preprint arXiv:2007.14527 (2020)

  39. Weinan, E., Yu, B.: The deep ritz method: a deep learning-based numerical algorithm for solving variational problems. Commun. Math. Stat. 6, 1–12 (2018)

    MathSciNet  MATH  Google Scholar 

  40. Yang, X., Golse, F., Huang, Z., Jin, S.: Numerical study of a domain decomposition method for a two-scale linear transport equation. Netw. Heterogeneous Media 1, 143 (2006)

    Article  MathSciNet  Google Scholar 

Download references


Y.L. thanks the US National Science Foundation for its support through the award DMS-2107934. L.W. and W.X. thank the National Science foundation for its support through the award DMS-1846854. The authors also acknowledge the Minnesota Supercomputing Institute (MSI) at the University of Minnesota for providing resources that contributed to the research results reported within this paper.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Wuzhe Xu.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Lu, Y., Wang, L. & Xu, W. Solving multiscale steady radiative transfer equation using neural networks with uniform stability. Res Math Sci 9, 45 (2022).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI:


Mathematics Subject Classification