Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
112 tokens/sec
GPT-4o
8 tokens/sec
Gemini 2.5 Pro Pro
47 tokens/sec
o3 Pro
5 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Solving multiscale dynamical systems by deep learning (2401.01220v2)

Published 2 Jan 2024 in math.NA and cs.NA

Abstract: Multiscale dynamical systems, modeled by high-dimensional stiff ordinary differential equations (ODEs) with wide-ranging characteristic timescales, arise across diverse fields of science and engineering, but their numerical solvers often encounter severe efficiency bottlenecks. This paper introduces a novel DeePODE method, which consists of an Evolutionary Monte Carlo Sampling method (EMCS) and an efficient end-to-end deep neural network (DNN) to predict multiscale dynamical systems. We validate this finding across dynamical systems from ecological systems to reactive flows, including a predator-prey model, a power system oscillation, a battery electrolyte thermal runaway, and turbulent reaction-diffusion systems with complex chemical kinetics. The method demonstrates robust generalization capabilities, allowing pre-trained DNN models to accurately predict the behavior in previously unseen scenarios, largely due to the delicately constructed dataset. While theoretical guarantees remain to be established, empirical evidence shows that DeePODE achieves the accuracy of implicit numerical schemes while maintaining the computational efficiency of explicit schemes. This work underscores the crucial relationship between training data distribution and neural network generalization performance. This work demonstrates the potential of deep learning approaches in modeling complex dynamical systems across scientific and engineering domains.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (44)
  1. Barlow R, Frank J (1998) Effects of turbulence on species mass fractions in methane/air jet flames in Symposium (International) on Combustion. (Elsevier), Vol. 27, pp. 1087–1095.
  2. McAdams HH, Shapiro L (1995) Circuit simulation of genetic networks. Science 269(5224):650–656.
  3. Volterra V (1931) Variations and fluctuations of the number of individuals in animal species living together. Animal ecology pp. 412–433.
  4. IEEE Transactions on Power Systems 32(2):1149–1160.
  5. Journal of computational neuroscience 27:369–390.
  6. Dissanayake M, Phan-Thien N (1994) Neural-network-based approximations for solving partial differential equations. communications in Numerical Methods in Engineering 10(3):195–201.
  7. Journal of Computational Physics 378:686–707.
  8. Zhang T, et al. (2022) A multi-scale sampling method for accurate and robust deep neural network to predict combustion chemical kinetics. Combustion and Flame 245:112319.
  9. Communications on Pure and Applied Mathematics 76(1):114–162.
  10. Milman VD, Schechtman G (2009) Asymptotic theory of finite dimensional normed spaces: Isoperimetric inequalities in riemannian manifolds. (Springer) Vol. 1200.
  11. Combustion Science and Technology 64(4-6):315–346.
  12. Pope SB (1997) Computationally efficient implementation of combustion chemistry using in situ adaptive tabulation.
  13. Combustion and Flame 106(4):406–427.
  14. Combustion and Flame 113(1-2):38–52.
  15. Choi Y, Chen JY (2005) Fast prediction of start-of-combustion in hcci with combined artificial neural networks and ignition delay model. Proceedings of the Combustion Institute 30(2):2711–2718.
  16. Sen BA, Menon S (2010) Linear eddy mixing based tabulation and artificial neural networks for large eddy simulations of turbulent flames. Combustion and Flame 157(1):62–74.
  17. Sinaei P, Tabejamaat S (2017) Large eddy simulation of methane diffusion jet flame with representation of chemical kinetics using artificial neural network. Proceedings of the Institution of Mechanical Engineers, Part E: Journal of Process Mechanical Engineering 231(2):147–163.
  18. (American Institute of Aeronautics and Astronautics Inc, AIAA).
  19. Owoyele O, Pal P (2022) Chemnode: A neural ordinary differential equations framework for efficient chemical kinetic solvers. Energy and AI 7:100118.
  20. Almeldein A, Van Dam N (2022) Accelerating chemical kinetics calculations with physics informed neural networks in Internal Combustion Engine Division Fall Technical Conference. (American Society of Mechanical Engineers), Vol. 86540, p. V001T06A007.
  21. Chaos: An Interdisciplinary Journal of Nonlinear Science 32(6):063107.
  22. Applications in Energy and Combustion Science 11:100077.
  23. Silver D, et al. (2016) Mastering the game of go with deep neural networks and tree search. nature 529(7587):484–489.
  24. Science 365(6457):eaaw1147.
  25. (IOP Publishing), Vol. 16, p. 65.
  26. Lu T, Law CK (2009) Toward accommodating realistic fuel chemistry in large-scale computations. Progress in Energy and Combustion Science 35(2):192–215.
  27. Combustion and Flame 226:467–477.
  28. Combustion and Flame 231:111493.
  29. Energy & Environmental Science 5(7):7797–7809.
  30. Seinfeld JH, Pandis SN (1998) From air pollution to climate change. Atmospheric chemistry and physics 1326.
  31. Maas U, Pope SB (1992) Simplifying chemical kinetics: intrinsic low-dimensional manifolds in composition space. Combustion and flame 88(3-4):239–264.
  32. Box GE, Cox DR (1964) An analysis of transformations. Journal of the Royal Statistical Society: Series B (Methodological) 26(2):211–243.
  33. International Conference on Neural Information Processing pp. 264–274.
  34. Communications in Computational Physics 28(5):1746–1767.
  35. SIAM Journal on Scientific Computing 43(3):A1607–A1624.
  36. pp. 1860–1863.
  37. Zirwes T, et al. (2020) Quasi-dns dataset of a piloted flame with inhomogeneous inlet conditions. Flow, Turbulence and Combustion 104(4):997–1027.
  38. Proceedings of the Combustion Institute 32(1):1253–1260.
  39. Proceedings of the Combustion Institute 38(2):2275–2283.
  40. Mao R, et al. (2022) Deepflame: A deep learning empowered open-source platform for reacting flow simulations. arXiv preprint arXiv:2210.07094.
  41. Wang Z, et al. (2022) A deep learning-based model reduction (deepmr) method for simplifying chemical kinetics.
  42. Combustion and Flame 135(1-2):185–190.
  43. Morev I, et al. (2022) Fast reactive flow simulations using analytical jacobian and dynamic load balancing in OpenFOAM. Physics of Fluids 34(2):021801.
  44. Computer Physics Communications p. 108073.
Citations (2)

Summary

We haven't generated a summary for this paper yet.