Abstract
This paper mainly introduces the parallel physics-informed neural networks (PPINNs) method with regularization strategies to solve the data-driven forward-inverse problems of the variable coefficient modified Korteweg-de Vries (VC-MKdV) equation. For the forward problem of the VC-MKdV equation, the authors use the traditional PINN method to obtain satisfactory data-driven soliton solutions and provide a detailed analysis of the impact of network width and depth on solving accuracy and speed. Furthermore, the author finds that the traditional PINN method outperforms the one with locally adaptive activation functions in solving the data-driven forward problems of the VC-MKdV equation. As for the data-driven inverse problem of the VC-MKdV equation, the author introduces a parallel neural networks to separately train the solution function and coefficient function, successfully addressing the function discovery problem of the VC-MKdV equation. To further enhance the network’s generalization ability and noise robustness, the author incorporates two regularization strategies into the PPINNs. An amount of numerical experimental data in this paper demonstrates that the PPINNs method can effectively address the function discovery problem of the VC-MKdV equation, and the inclusion of appropriate regularization strategies in the PPINNs can improves its performance.
Similar content being viewed by others
References
Carslaw H S, Introduction to the Mathematical Theory of the Conduction of Heat in Solids, Macmillan and Company, London, 1906.
Batchelor C K, An Introduction to Fluid Dynamics, Cambridge University Press, Cambridge, 1967.
Cohen-Tannoudji C, Diu B, and Laloë F, Quantum Mechanics: Concepts and Applications, John Wiley and Sons, New York, 2019.
Keener J P and Sneyd J, Mathematical Physiology 1: Cellular Physiology, Springer Science and Business Media, New York, 2009.
Djordjevic V and Redekopp L, The fission and disintegration of internal solitary waves moving over two-dimensional topography, J. Phys. Oceanogr., 1978, 8): 1016–1024.
Porsezian K, Generalized x-dependent modified Korteweg-de Vries equation: Painlevé analysis, Bäckhand transformation and soliton solutions, Phys. Lett. A, 1996, 221): 163–166.
Yan Z Y, The modified KdV equation with variable coefficients: Exact uni/bivariable travelling wave-like solutions, Appl. Math. Comput., 2008, 203): 106–112.
Zhang Y, Li J B, and Lü Y N, The exact solution and integrable properties to the variable-coefficient modified Korteweg-de Vries equation, Ann. Phys., 2008, 323): 3059–3064.
Vaneeva O, Lie symmetries and exact solutions of variable coefficient mKdV equations: An equivalence based approach, Commun. Nonlinear Sci. Numer. Simulat., 2012, 17): 611–618.
Houria T, Taha T R, and Wazwaz A M, Solitary wave solutions for a generalized KdV-mKdV equation with variable coefficients, Math. Comput. Simulat., 2010, 80): 1867–1873.
Gai X L, Gao Y T, Wang L, et al., Painlevé property, Lax pair and Darboux transformation of the variable-coefficient modified Kortweg-de Vries model in fluid-filled elastic tubes, Commun. Nonlinear Sci. Numer. Simulat., 2011, 16): 1776–1782.
Wang L, Gao Y T, and Qi F H, N-fold Darboux transformation and double Wronskian-typed solitonic structures for a variable-coefficient modified Kortweg-de Vries equation, Ann. Phys., 2012, 327): 1974–1988.
Lin Y, Li C Z, and He J S, Nonsingular positon solutions of a variable-coefficient modified KdV equation, Open Journal Applied Sciences, 2013, 3): 102–105.
Yang D Y, Tian B, Qu Q X, et al., Lax pair, conservation laws, Darboux transformation and localized waves of a variable-coefficient coupled Hirota system in an inhomogeneous optical fiber, Chaos, Soliton. Fract., 2021, 150): 110487.
Mou D S and Dai C Q, Nondegenerate solitons and collision dynamics of the variable-coefficient coupled higher-order nonlinear Schrödinger model via the Hirota method, Appl. Math. Lett., 2022, 133): 108230.
Liu W J, Yang C Y, Liu M L, et al., Effect of high-order dispersion on three-soliton interactions for the variable-coefficients Hirota equation, Phys. Rev. E, 2017, 96): 042201.
Zhou H J and Chen Y, High-order soliton solutions and their dynamics in the inhomogeneous variable coefficients Hirota equation, Commun. Nonlinear Sci. Numer. Simul., 2023, 120): 107149.
Tian B, Gao Y T, and Zhu H W, Variable-coefficient higher-order nonlinear Schrödinger model in optical fibers: Variable-coefficient bilinear form, Bäcklund transformation, brightons and symbolic computation, Phys. Lett. A, 2007, 366(3): 223–229.
Fan E G, Auto-Backlund transformation and similarity reductions for general variable coefficient KdV equations, Phys. Lett. A, 2002, 294(1): 26–30.
Hornik K, Stinchcombe M, and White H, Universal approximation of an unknown mapping and its derivatives using multilayer feedforward networks, Neural Netw., 1990, 3): 551–560.
Zhang R F, Li M C, Cherraf A, et al., The interference wave and the bright and dark soliton for two integro-differential equation by using BNNM, Nonlinear Dyn., 2023, 111(9): 8637–8646.
Zhang R F, Li M C, Gan J Y, et al., Novel trial functions and rogue waves of generalized breaking soliton equation via bilinear neural network method, Chaos, Soliton. Fract., 2022, 154): 111692.
Zhang R F, Li M C, Albishari M, et al., Generalized lump solutions, classical lump solutions and rogue waves of the (2+1)-dimensional Caudrey-Dodd-Gibbon-Kotera-Sawada-like equation, Appl. Math. Comput., 2021, 403): 126201.
Raissi M, Perdikaris P, and 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., 2019, 378): 686–707.
Jagtap A D, Kawaguchi K, and Karniadakis G E, Adaptive activation functions accelerate convergence in deep and physics-informed neural networks, J. Comput. Phys., 2020, 404: 109136.
Jagtap A D, Kawaguchi K, and Karniadakis G E, Locally adaptive activation functions with slope recovery for deep and physics-informed neural networks, Proc. R. Soc. A., 2020, 476): 20200334.
Li J and Chen Y, Solving second-order nonlinear evolution partial differential equations using deep learning, Commun. Theor. Phys., 2020, 72: 105005.
Li J and Chen Y, A deep learning method for solving third-order nonlinear evolution equations, Commun. Theor. Phys., 2020, 72): 115003.
Pu J C, Li J, and Chen Y, Soliton, breather and rogue wave solutions for solving the nonlinear Schrödinger equation using a deep learning method with physical constraints, Chin. Phys. B., 2021, 30): 060202.
Peng W Q, Pu J C, and Chen Y, PINN deep learning for the Chen-Lee-Liu equation: Rogue wave on the periodic background, Commun. Nonlinear Sci. Numer. Simul., 2022, 105): 106067.
Miao Z W and Chen Y, Physics-informed neural network method in high-dimensional integrable systems, Mode. Phys. Lett. B., 2021, 36): 2150531.
Lin S N and Chen Y, A two-stage physics-informed neural network method based on conserved quantities and applications in localized wave solutions, J. Comput. Phys., 2022, 457: 111053.
Lin S N and Chen Y, Physics-informed neural network methods based on Miura transformations and discovery of new localized wave solutions, Physica D, 2023, 445): 133629.
Fang Y, Wu G Z, Wang Y Y, et al., Data-driven femtosecond optical soliton excitations and parameters discovery of the high-order NLSE using the PINN, Nonlinear Dyn., 2021, 105): 603–616.
Wang L and Yan Z Y, Data-driven rogue waves and parameter discovery in the defocusing nonlinear Schrödinger equation with a potential using the PINN deep learning, Phys. Lett. A., 2021, 404): 127408.
Wu G Z, Fang Y, Wang Y Y, et al., Predicting the dynamic process and model parameters of the vector optical solitons in birefringent fibers via the modified PINN, Chaos, Soliton. Fract., 2021, 152): 111393.
Mo Y F, Ling L M, and Zeng D L, Data-driven vector soliton solutions of coupled nonlinear Schrödinger equation using a deep learning algorithm, Phys. Lett. A., 2022, 421): 127739.
Li J H, Chen J C, and Li B, Gradient-optimized physics-informed neural networks (GOPINNs): A deep learning method for solving the complex modified KdV equation, Nonlinear Dyn., 2022, 107): 781–792.
Zhou Z J and Yan Z Y, Solving forward and inverse problems of the logarithmic nonlinear Schrödinger equation with-symmetric harmonic potential via deep learning, Phys. Lett. A., 2021, 387): 127010.
Miao Z W and Chen Y, VC-PINN: Variable coefficient physical information neural network for forward and inverse PDE problems with variable coefficient, Physica D, 2023, 456): 133945.
Lin S N and Chen Y, Gradient-enhanced physics-informed neural networks based on transfer learning for inverse problems of the variable coefficient differential equations, 2023, arXiv: 2305.08310v1.
Zhou H J, Pu J C, and Chen Y, Data-driven forwardd-inverse problems for the variable coefficients Hirota equation using deep learning method, Nonlinear Dyn., 2023, 111): 146671–4693.
Kingma D P and Ba J L, Adam: A method for stochastic optimization, 2014, arXiv: 1412.6980.
Liu D C and Nocedal J, On the limited memory BFGS method for large scale optimization, Math Program., 1989, 45): 503–528.
Acknowledgements
The authors are profoundly grateful to Yong Chen and Zhengwu Miao for their invaluable remarks and proposals.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
The authors declare no conflict of interest.
Rights and permissions
About this article
Cite this article
Zhou, H. Parallel Physics-Informed Neural Networks Method with Regularization Strategies for the Forward-Inverse Problems of the Variable Coefficient Modified KdV Equation. J Syst Sci Complex 37, 511–544 (2024). https://doi.org/10.1007/s11424-024-3467-7
Received:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11424-024-3467-7