Abstract
Physics-informed neural networks (PINNs) have been an effective tool for approximating the map** between points in the spatio-temporal domain and solutions of partial differential equations (PDEs). However, there are still some challenges in dealing with the nonlinear characteristics and complexity of the Navier–Stokes (N–S) equations. In this paper, the improved adaptive weighting PINNs based on the Gaussian likelihood estimation are applied to solve the N–S equations. The weights of the different loss items are allocated adaptively by the maximum likelihood estimation. The improved network structure has been designed with considering both the global and local information, making it easier to capture the part of PDEs solution with drastic changes. A combinational method of the numerical differentiation (ND) and the automatic differentiation (AD) is proposed to compute the differential operators, with the improved computational efficiency. The derivative operation of the convection and pressure-gradient terms was carried out using the combined method in solving the incompressible N–S equations. The results show that the effectiveness and training efficiency of this method are better than PINNs.
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig1_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig2_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig3_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig4_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig5_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig6_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig7_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig8_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig9_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig10_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig11_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig12_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig13_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig14_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig15_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig16_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig17_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig18_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig19_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig20_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig21_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11071-024-09856-6/MediaObjects/11071_2024_9856_Fig22_HTML.png)
Similar content being viewed by others
Data availability
The code can be found at https://github.com/upwaj/aw-in-md-PINNs.
References
Girault, V., Raviart, P.-A.: Finite element methods for Navier–Stokes equations: theory and algorithms, vol. 5. Springer Science & Business Media, Cham (2012)
Coelho, P.J., Pereira, J.C.F.: Finite volume computation of the turbulent flow over a hill employing 2D or 3D non-orthogonal collocated grid systems. Int. J. Numer. Methods Fluids 14(4), 423–441 (1992)
Weinan, E., Liu, J.-G.: Finite difference methods for 3D viscous incompressible flows in the vorticity-vector potential formulation on nonstaggered grids. J. Comput. Phys. 138(1), 57–82 (1997)
Ranade, R., Hill, C., Pathak, J.: DiscretizationNet: a machine-learning based solver for Navier–Stokes equations using finite volume discretization. Comput. Methods Appl. Mech. Eng. 378, 113722 (2021)
Wang, Y., Lai, C.-Y.: Multi-stage neural networks: function approximator of machine precision. J. Comput. Phys. 504, 112865 (2024)
Li, X., Liu, Y., Liu, Z.: Physics-informed neural network based on a new adaptive gradient descent algorithm for solving partial differential equations of flow problems. Phys. Fluids 35(6), 063608 (2023)
Eivazi, H., Tahani, M., Schlatter, P., Vinuesa, R.: Physics-informed neural networks for solving Reynolds-averaged Navier–Stokes equations. Phys. Fluids 34(7), 075117 (2022)
**ao, M.-J., Teng-Chao, Y., Zhang, Y.-S., Yong, H.: Physics-informed neural networks for the Reynolds-averaged Navier–Stokes modeling of Rayleigh–Taylor turbulent mixing. Comput. Fluids 266, 106025 (2023)
Azizzadenesheli, K., Kovachki, N., Li, Z., Liu-Schiaffini, M., Kossaifi, J., Anandkumar, A.: Neural operators for accelerating scientific simulations and design. Nat. Rev. Phys. 6, 1–9 (2024)
Lee, H., Kang, I.S.: Neural algorithm for solving differential equations. J. Comput. Phys. 91(1), 110–131 (1990)
Lagaris, I.E., Likas, A., Fotiadis, D.I.: Artificial neural networks for solving ordinary and partial differential equations. IEEE Trans. Neural Netw. 9(5), 987–1000 (1998)
Weinan, E., Bing, Y.: The deep Ritz method: a deep learning-based numerical algorithm for solving variational problems. Commun. Math. Stat. 6, 1–12 (2018)
Chen, M., Niu, R., Zheng, W.: Adaptive multi-scale neural network with resnet blocks for solving partial differential equations. Nonlinear Dyn. 111, 6499–6518 (2022)
Gao, R., Wei, H., Fei, J., Hongyu, W.: Boussinesq equation solved by the physics-informed neural networks. Nonlinear Dyn. 111, 15279–15291 (2023)
Zhang, T., Hui, X., Guo, L., Feng, X.: A non-intrusive neural network model order reduction algorithm for parameterized parabolic PDEs. Comput. Math. Appl. 119, 59–67 (2022)
Hou, J., Li, Y., Ying, S.: Enhancing PINNs for solving PDEs via adaptive collocation point movement and adaptive loss weighting. Nonlinear Dyn. 111, 15233–15261 (2023)
Wang, H., Zou, B., Jian, S., Wang, D.: Variational methods and deep Ritz method for active elastic solids. Soft Matter 18(7), 6015–6031 (2022)
Gao, Z., Yan, L., Zhou, T.: Failure-informed adaptive sampling for PINNs. SIAM J. Sci. Comput. 45, A1971–A1994 (2023)
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)
Karniadakis, G.E., Kevrekidis, I.G., Lu, L., Perdikaris, P., Wang, S., Yang, L.: Physics-informed machine learning. Nat. Rev. Phys. 3(6), 422–440 (2021)
Lu, L., Meng, X., Mao, Z., Karniadakis, G.E.: DeepXDE: a deep learning library for solving differential equations. SIAM Rev. 63(1), 208–228 (2021)
Chiu, P.-H., Wong, J.C., Ooi, C., Dao, M.H., Ong, Y.-S.: CAN-PINN: a fast physics-informed neural network based on coupled-automatic-numerical differentiation method. Comput. Methods Appl. Mech. Eng. 395, 114909 (2022)
Patel, R.G., Manickam, I., Trask, N.A., Wood, M.A., Lee, M., Tomas, Ignacio, Cyr, Eric C.: Thermodynamically consistent physics-informed neural networks for hyperbolic systems. J. Comput. Phys. 449, 110754 (2022)
Jagtap, A.D., Kharazmi, E., Karniadakis, G.E.: Conservative physics-informed neural networks on discrete domains for conservation laws: applications to forward and inverse problems. Comput. Methods Appl. Mech. Eng. 365, 113028 (2020)
Jagtap, A.D., Karniadakis, G.E.: Extended physics-informed neural networks (XPINNs): a generalized space-time domain decomposition based deep learning framework for nonlinear partial differential equations. Commun. Comput. Phys. 28, 2002–2041 (2020)
Wei, W., Feng, X., Hui, X.: Improved deep neural networks with domain decomposition in solving partial differential equations. J. Sci. Comput. 93, 1–34 (2022)
Wang, S., Sankaran, S., Wang, H., Perdikaris, P.: An expert’s guide to training physics-informed neural networks (2023). ar**v preprint: ar**v: 2308.08468
McClenny, L.D., Braga-Neto, U.M.: Self-adaptive physics-informed neural networks. J. Comput. Phys. 474, 111722 (2023)
Tang, S., Feng, X., Wei, W., Hui, X.: Physics-informed neural networks combined with polynomial interpolation to solve nonlinear partial differential equations. Comput. Math. Appl. 132, 48–62 (2023)
Peng, P., Pan, J., Hui, X., Feng, X.: RPINNs: rectified-physics informed neural networks for solving stationary partial differential equations. Comput. Fluids 245, 105583 (2022)
Chenxi, W., Zhu, M., Tan, Q., Kartha, Y., Lu, L.: A comprehensive study of non-adaptive and residual-based adaptive sampling for physics-informed neural networks. Comput. Methods Appl. Mech. Eng. 403, 115671 (2023)
Bai, Y., Chaolu, T., Bilige, S.: Solving Huxley equation using an improved PINN method. Nonlinear Dyn. 105, 3439–3450 (2021)
**, X., Cai, S., Li, H., Karniadakis, G.E.: NSFnets (Navier–Stokes flow nets): physics-informed neural networks for the incompressible Navier–Stokes equations. J. Comput. Phys (2021). https://doi.org/10.1016/j.jcp.2020.109951
Rao, C., Sun, H., Liu, Y.: Physics-informed deep learning for incompressible laminar flows. Theor. Appl. Mech. Lett. 10(3), 207–212 (2020)
Dwivedi, B.S.V., Parashar, N.: Distributed learning machines for solving forward and inverse problems in partial differential equations. Neurocomputing 420, 299–316 (2021)
Gao, H., Sun, L., Wang, J.-X.: PhyGeoNet: physics-informed geometry-adaptive convolutional neural networks for solving parameterized steady-state PDEs on irregular domain. J. Comput. Phys. 428, 110079 (2021)
Sitzmann, V., Martel, J., Bergman, A., Lindell, D., Wetzstein, G.: Implicit neural representations with periodic activation functions. Adv. Neural Inf. Process. Syst. 33, 7462–7473 (2020)
Mattey, R., Ghosh, S.: A novel sequential method to train physics informed neural networks for Allen–Cahn and Cahn–Hilliard equations. Comput. Methods Appl. Mech. Eng. 390, 114474 (2022)
Stein, M.: Large sample properties of simulations using Latin hypercube sampling. Technometrics 29(2), 143–151 (1987)
Liu, D.C., Nocedal, J.: On the limited memory BFGS method for large scale optimization. Math. Program. 45(1–3), 503–528 (1989)
Kendall, A., Gal, Y., Cipolla, R.: Multi-task learning using uncertainty to weigh losses for scene geometry and semantics. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 7482–7491 (2018)
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, Ł, Polosukhin, I.: Attention is all you need. Adv. Neural Inf. Process. Syst. 30, 1 (2017)
Cho, K., Merrienboer, B., Gulcehre, C., Bougares, F., Schwenk, H., Bengio, Y.: Learning phrase representations using RNN encoder-decoder for statistical machine translation. In: Conference on Empirical Methods in Natural Language Processing (EMNLP 2014) (2014)
Abadi, M., Barham, P., Chen, J., Chen, Z., Davis, A., Dean, J., Devin, M., Ghemawat, S., Irving, G., Isard, M.: Tensorflow: a system for large-scale machine learning on heterogeneous distributed systems (2016). ar**v: 1603.04467
**ao, F., Akoh, R., Ii, S.: Unified formulation for compressible and incompressible flows by using multi-integrated moments ii: multi-dimensional version for compressible and incompressible flows. J. Comput. Phys. 213(1), 31–56 (2006)
Sheu, T.W.H., Chiu, P.H.: A divergence-free-condition compensated method for incompressible Navier–Stokes equations. Comput. Methods Appl. Mech. Eng. 196(45–48), 4479–4494 (2007)
Ethier, C.R., Steinman, D.A.: Exact fully 3D Navier–Stokes solutions for benchmarking. Int. J. Numer. Methods Fluids 19(5), 369–375 (1994)
Ghia, U.K.N.G., Ghia, K.N., Shin, C.T.: High-Re solutions for incompressible flow using the Navier–Stokes equations and a multigrid method. J. Comput. Phys. 48(3), 387–411 (1982)
Williamson, C.H.K.: Vortex dynamics in the cylinder wake. Ann. Rev. Fluid Mech. 28(1), 477–539 (1996)
Funding
This work is supported by the National Natural Science Foundation of China (Nos. 12361090, 12071406 and 12271465), the Natural Science Foundation of **njiang Province (Nos. 2023D01C164, 6142A05230203, 2022D01D32), and the Tianshan Talent Training Program (No. 2023TSYCQNTJ0015 and No. 2022TSYCTD0019).
Author information
Authors and Affiliations
Contributions
Jie Wang: investigation, software, validation, writing—original draft. Xufeng **ao: conceptualization, methodology, formal analysis, software, writing—review and editing. **nlong Feng: formal analysis, writing—review and editing. Hui Xu: formal analysis, writing—review and editing.
Corresponding author
Ethics declarations
Conflict of interest
No conflict of interest exists in the submission of this manuscript, and the manuscript is approved by all authors for publication.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Wang, J., **ao, X., Feng, X. et al. An improved physics-informed neural network with adaptive weighting and mixed differentiation for solving the incompressible Navier–Stokes equations. Nonlinear Dyn (2024). https://doi.org/10.1007/s11071-024-09856-6
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s11071-024-09856-6