A residual-based adaptive refinement physics-informed neural networks (RAR-PINNs) method for fifth-order KdV equation

doi:10.1088/1674-1056/ae1458

Abstract In the field of nonlinear partial differential equations (PDEs), the fifth-order Korteweg-De Vries (KdV) equation serves as a fundamental model with significant physical implications, extending the classical KdV framework through the incorporation of high-order spatial derivatives to capture strong dispersion effects. However, the inherent nonlinearity and complexity of this PDE present substantial challenges for obtaining accurate numerical solutions. To address these issues, this paper proposes a residual-based adaptive refinement physics-informed neural networks (RAR-PINNs) method. This approach synergizes the nonlinear approximation capability of PINNs with a residual-driven adaptive sampling strategy. By dynamically redistributing training points according to the magnitude of the PDE residuals, RAR-PINNs effectively concentrate computational resources on ``critical regions'', such as soliton peaks and high-gradient zones, where errors are predominant. Furthermore, we construct a composite physics-informed loss function that incorporates initial and boundary conditions, PDE residuals, and, in an enhanced variant, energy conservation laws, to further improve solution fidelity. Numerical experiments on two variants of the fifth-order KdV equation demonstrate that RAR-PINNs significantly outperform conventional PINNs in terms of both accuracy (reducing relative errors by one to two orders of magnitude) and computational efficiency. The conservation-law-enhanced version of the model yields even higher precision, underscoring the efficacy and robustness of the proposed method. This study establishes a powerful deep learning framework for tackling complex PDEs with sharp or singular solution structures.

Keywords: data-driven PINNs residual-based adaptive refinement fifth-order KdV equation

Received: 17 September 2025
Revised: 15 October 2025
Accepted manuscript online: 17 October 2025

PACS:	02.30.Ik	(Integrable systems)
	02.30.Jr	(Partial differential equations)
	05.45.-a	(Nonlinear dynamics and chaos)

Fund: This work was supported by the National Natural Science Foundation of China (Grant Nos. 12175111 and 12235007), and the K. C. Wong Magna Fund in Ningbo University.

Corresponding Authors: Biao Li
E-mail: libiao@nbu.edu.cn

Cite this article:

Shi-Fang Tian(田十方), Ya-Xuan Yu(于亚璇), and Biao Li(李彪) A residual-based adaptive refinement physics-informed neural networks (RAR-PINNs) method for fifth-order KdV equation 2026 Chin. Phys. B 35 050201

[1] Korteweg D J and de Vries G 1895 The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science 39 422
[2] Wang M L and Li M 2025 Acta Phys. Sin. 74 150201 (in Chinese)
[3] Wazwaz A M 2016 Journal of Taibah University for Science 10 469
[4] Zhang F, Hu Y R, Xin X P and Liu H Z 2022 Computational and Applied Mathematics 41 139
[5] Liu N and Guo B 2021 Anal. Appl. 19 575
[6] Niu Z J and Wang Z G 2021 Journal of Applied Analysis & Computation 11 2866
[7] Wu W J 1978 Scientia Sinica 21 159
[8] Raissi M, Perdikaris P and Karniadakis G E 2019 J. Comput. Phys. 378 686
[9] Jagtap A D, Mao Z P, Adams N and Karniadakis G E 2022 J. Comput. Phys. 466 111402
[10] Pang G F, Lu L and Karniadakis G E 2022 SIAM J. Sci. Comput. 41 2603
[11] Peng W Q and Chen Y 2022 Physica D 435 133274
[12] Pu J C and Chen Y 2022 Chaos, Solitons and Fractals 160 112182
[13] Wang L and Yan Z Y 2021 Phys. Lett. A 404 127408
[14] Wang L and Yan Z Y 2022 Phys. Lett. A 450 128373
[15] Fang Y, Wu G Z, Wang Y Y and Dai C Q 2021 Nonlinear Dyn. 105 603
[16] Wen X K, Jiang J H, Liu W and Dai C Q 2023 Nonlinear Dyn. 111 13343
[17] Guo J, Wang H F, Gu S L and Hou C P 2024 Chin. Phys. B 33 050701
[18] Zhou H S, Zhang Z Y, Liu M W, Qi F H and Liu W J 2025 Chin. Phys. Lett. 42 070001
[19] Li J H and Li B 2022 Nonlinear Dyn. 107 781
[20] Tian S F and Li B 2023 Acta Phys. Sin. 72 100202 (in Chinese)
[21] Li J H and Li B 2022 Chaos, Solitons and Fractals 164 112712
[22] Tian S F, Zhang Z and Li B 2024 Chin. Phys. Lett. 41 030201
[23] Lu L, Meng X M, Mao Z P and Karniadakis G E 2021 SIAM Review 63 10
[24] Xu S Y, Zhou Q and Liu W 2023 Nonlinear Dyn. 111 18401
[25] Sun F W and Wang M 2013 Mathematics in Practice and Theory 43 256
[26] LiWT, Chen Y P, Zhang Z, Yang X Y and Li B 2025 Chin. Phys. Lett. 42 010202

[1]	Physics-informed neural network for material identification via distortion-robust polychromatic x-ray attenuation correction in photon-counting detectors Xin Yan(闫欣), Jie Zhang(张杰), Kai He(何凯), Yiheng Liu(刘毅恒), Yuetong Zhao(赵悦彤), Gang Wang(王刚), Xinlong Chang(常新龙), and Youwei Zhang(张有为). Chin. Phys. B, 2026, 35(5): 050702.
[2]	Detecting physical laws from data of stochastic dynamical systems perturbed by non-Gaussian α-stable Lévy noise Linghongzhi Lu(陆凌弘志), Yang Li(李扬), and Xianbin Liu(刘先斌). Chin. Phys. B, 2023, 32(5): 050501.
[3]	Data-driven modeling of a four-dimensional stochastic projectile system Yong Huang(黄勇) and Yang Li(李扬). Chin. Phys. B, 2022, 31(7): 070501.
[4]	Sparse identification method of extracting hybrid energy harvesting system from observed data Ya-Hui Sun(孙亚辉), Yuan-Hui Zeng(曾远辉), and Yong-Ge Yang(杨勇歌). Chin. Phys. B, 2022, 31(12): 120203.
[5]	Line positions, intensities, and Einstein A coefficients for 3-0 band of ¹²C¹⁶O: A spectroscopy learning method Zhi-Xiang Fan(范志祥), Zhi-Zhang Ni(倪志樟), Jie-Jie He(贺洁洁), Yi-Fan Wang(王一凡), Qun-Chao Fan(樊群超), Jia Fu(付佳), Yong-Gen Xu(徐勇根), Hui-Dong Li(李会东), Jie Ma(马杰), and Feng Xie(谢锋). Chin. Phys. B, 2021, 30(12): 123301.
[6]	Painlevé integrability of generalized fifth-order KdV equation with variable coefficients: Exact solutions and their interactions Xu Gui-Qiong (徐桂琼). Chin. Phys. B, 2013, 22(5): 050203.
[7]	Multi-symplectic method for generalized fifth-order KdV equation Hu Wei-Peng (胡伟鹏), Deng Zi-Chen (邓子辰). Chin. Phys. B, 2008, 17(11): 3923-3929.
[8]	The novel multi-solitary wave solution to the fifth-order KdV equation Zhang Yi (张翼), Chen Deng-Yuan (陈登远). Chin. Phys. B, 2004, 13(10): 1606-1610.

No Suggested Reading articles found!

Viewed

Full text

Abstract

Cited

Metrics
Related Articles

A residual-based adaptive refinement physics-informed neural networks (RAR-PINNs) method for fifth-order KdV equation

Cite this article:

Online attention