Gradient-enhanced PINN with residual unit for studying forward-inverse problems of variable coefficient equations
Physics-informed neural network (PINN) is a powerful emerging method for studying forward-inverse problems of partial differential equations (PDEs), even from limited sample data. Variable coefficient PDEs, which model real-world phenomena, are of considerable physical significance and research valu...
Uloženo v:
| Vydáno v: | Physica. D Ročník 481; s. 134764 |
|---|---|
| Hlavní autoři: | , |
| Médium: | Journal Article |
| Jazyk: | angličtina |
| Vydáno: |
Elsevier B.V
01.11.2025
|
| Témata: | |
| ISSN: | 0167-2789 |
| On-line přístup: | Získat plný text |
| Tagy: |
Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
|
| Abstract | Physics-informed neural network (PINN) is a powerful emerging method for studying forward-inverse problems of partial differential equations (PDEs), even from limited sample data. Variable coefficient PDEs, which model real-world phenomena, are of considerable physical significance and research value. This study proposes a gradient-enhanced PINN with residual unit (R-gPINN) method to solve the data-driven solution and function discovery for variable coefficient PDEs. On the one hand, the proposed method incorporates residual units into the neural networks to mitigate gradient vanishing and network degradation, unify linear and nonlinear coefficient problem. We present two types of residual unit structures in this work to offer more flexible solutions in problem-solving. On the other hand, by including gradient terms of variable coefficients, the method penalizes collocation points that fail to satisfy physical properties. This enhancement improves the network’s adherence to physical constraints and aligns the prediction function more closely with the objective function. Numerical experiments including solve the forward-inverse problems of variable coefficient Burgers equation, variable coefficient KdV equation, variable coefficient Sine–Gordon equation, and high-dimensional variable coefficient Kadomtsev–Petviashvili equation. The results show that using R-gPINN method can greatly improve the accuracy of predict solution and predict variable coefficient in solving variable coefficient equations. |
|---|---|
| AbstractList | Physics-informed neural network (PINN) is a powerful emerging method for studying forward-inverse problems of partial differential equations (PDEs), even from limited sample data. Variable coefficient PDEs, which model real-world phenomena, are of considerable physical significance and research value. This study proposes a gradient-enhanced PINN with residual unit (R-gPINN) method to solve the data-driven solution and function discovery for variable coefficient PDEs. On the one hand, the proposed method incorporates residual units into the neural networks to mitigate gradient vanishing and network degradation, unify linear and nonlinear coefficient problem. We present two types of residual unit structures in this work to offer more flexible solutions in problem-solving. On the other hand, by including gradient terms of variable coefficients, the method penalizes collocation points that fail to satisfy physical properties. This enhancement improves the network’s adherence to physical constraints and aligns the prediction function more closely with the objective function. Numerical experiments including solve the forward-inverse problems of variable coefficient Burgers equation, variable coefficient KdV equation, variable coefficient Sine–Gordon equation, and high-dimensional variable coefficient Kadomtsev–Petviashvili equation. The results show that using R-gPINN method can greatly improve the accuracy of predict solution and predict variable coefficient in solving variable coefficient equations. |
| ArticleNumber | 134764 |
| Author | Zhou, Hui-Juan Chen, Yong |
| Author_xml | – sequence: 1 givenname: Hui-Juan surname: Zhou fullname: Zhou, Hui-Juan organization: School of Science, Shanghai Maritime University, Shanghai, 201306, PR China – sequence: 2 givenname: Yong orcidid: 0000-0002-6008-6542 surname: Chen fullname: Chen, Yong email: ychen@sei.ecnu.edu.cn organization: School of Mathematical Sciences, Key Laboratory of MEA (Ministry of Education) and Shanghai Key Laboratory of PMMP, East China Normal University, Shanghai, 200241, China |
| BookMark | eNp9kEFPAjEUhHvAREB_gZf-gV3b7m539-DBEEUSgh703JT2VUqghbYL4d-7K549vUzezGTyTdDIeQcIPVCSU0L54zY_bC5R54ywKqdFWfNyhMb9p85Y3bS3aBLjlhBC66Ieo-M8SG3BpQzcRjoFGn8sVit8tmmDA0SrO7nDnbMJGx9wTJ2-WPc9iLMMOrPuBCECPgS_3sE-Ym_wSQYre4WVB2OsGuoxHDuZrHfxDt0YuYtw_3en6Ov15XP2li3f54vZ8zJTrKpSxriRirJCl6wFxrkma6pJq5lhQBtWKWLKVhrerGmraFnWtJWEN5z1Xq6Kupii4tqrgo8xgBGHYPcyXAQlYiAltuKXlBhIiSupPvV0TUE_7WQhiDjs77nYACoJ7e2_-R-6iHio |
| Cites_doi | 10.1016/j.physleta.2021.127739 10.1016/j.physleta.2022.128536 10.1016/j.physd.2023.133729 10.1143/JPSJ.30.272 10.1103/PhysRevE.108.045303 10.1002/sapm1987762133 10.1016/0021-9991(83)90085-2 10.1016/j.chaos.2021.111393 10.1016/j.physd.2022.133629 10.1016/0893-6080(89)90020-8 10.1016/j.cma.2022.114823 10.1016/j.ijleo.2019.01.018 10.1007/s11071-023-08641-1 10.1002/mma.5982 10.1017/S0022112078001214 10.1016/j.physd.2023.133851 10.1016/j.jcp.2024.113090 10.1016/j.jcp.2018.10.045 10.1002/sapm19898011 10.1016/j.physd.2023.133945 10.1140/epjd/e2011-20518-0 10.1016/S0375-9601(02)00033-6 10.1103/PhysRevLett.40.233 10.1016/j.physleta.2021.127408 10.1007/s11071-023-08712-3 10.1007/s11424-024-3467-7 10.1016/S0022-247X(02)00445-6 10.1016/0045-7825(92)90088-2 10.1063/1.1693097 10.1038/s42254-021-00314-5 10.1090/qam/42889 10.1016/0375-9601(75)90124-3 |
| ContentType | Journal Article |
| Copyright | 2025 |
| Copyright_xml | – notice: 2025 |
| DBID | AAYXX CITATION |
| DOI | 10.1016/j.physd.2025.134764 |
| DatabaseName | CrossRef |
| DatabaseTitle | CrossRef |
| DatabaseTitleList | |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Engineering Physics |
| ExternalDocumentID | 10_1016_j_physd_2025_134764 S0167278925002416 |
| GroupedDBID | --K --M -~X .~1 0R~ 1B1 1RT 1~. 1~5 29O 4.4 457 4G. 5VS 7-5 71M 8P~ 9JN AAEDT AAEDW AAIKJ AAKOC AALRI AAOAW AAQFI AAQXK AATTM AAXKI AAXUO AAYWO ABAOU ABFNM ABMAC ABNEU ABWVN ABXDB ACDAQ ACFVG ACGFS ACLOT ACNCT ACNNM ACRLP ACRPL ADBBV ADEZE ADGUI ADIYS ADMUD ADNMO ADVLN AEBSH AEIPS AEKER AFFNX AFJKZ AFTJW AGHFR AGQPQ AGUBO AGYEJ AHHHB AIEXJ AIGVJ AIIUN AIKHN AITUG AIVDX ALMA_UNASSIGNED_HOLDINGS AMRAJ ANKPU APXCP ARUGR ASPBG AVWKF AXJTR AZFZN BBWZM BKOJK BLXMC EBS EFJIC EFKBS EFLBG EJD EO8 EO9 EP2 EP3 F5P FDB FEDTE FGOYB FIRID FNPLU FYGXN G-Q GBLVA HMV HVGLF HZ~ H~9 IHE J1W K-O KOM M38 M41 MHUIS MO0 MVM N9A NDZJH O-L O9- OAUVE OGIMB OZT P-8 P-9 P2P PC. Q38 R2- RNS ROL RPZ SDF SDG SDP SES SEW SPC SPCBC SPD SPG SSQ SSW SSZ T5K TN5 TWZ WUQ XJT XPP YNT YYP ~02 ~G- ~HD 9DU AAYXX CITATION |
| ID | FETCH-LOGICAL-c255t-26fac123d429e266d0b1d09d2f2e1825c0f49af68b19c144719a068622666c373 |
| ISICitedReferencesCount | 1 |
| ISICitedReferencesURI | http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001516751600001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| ISSN | 0167-2789 |
| IngestDate | Sat Nov 29 07:37:03 EST 2025 Sat Sep 27 17:12:47 EDT 2025 |
| IsPeerReviewed | true |
| IsScholarly | true |
| Keywords | Physics-informed neural network Function discovery Variable coefficient equation Residual network Data-driven solution |
| Language | English |
| LinkModel | OpenURL |
| MergedId | FETCHMERGED-LOGICAL-c255t-26fac123d429e266d0b1d09d2f2e1825c0f49af68b19c144719a068622666c373 |
| ORCID | 0000-0002-6008-6542 |
| ParticipantIDs | crossref_primary_10_1016_j_physd_2025_134764 elsevier_sciencedirect_doi_10_1016_j_physd_2025_134764 |
| PublicationCentury | 2000 |
| PublicationDate | November 2025 2025-11-00 |
| PublicationDateYYYYMMDD | 2025-11-01 |
| PublicationDate_xml | – month: 11 year: 2025 text: November 2025 |
| PublicationDecade | 2020 |
| PublicationTitle | Physica. D |
| PublicationYear | 2025 |
| Publisher | Elsevier B.V |
| Publisher_xml | – name: Elsevier B.V |
| References | Ott, Sudan (b9) 1970; 13 Zhou (b26) 2024; 37 Güngör, Winternitz (b46) 2002; 276 Song, Yan (b29) 2023; 448 Moghimi, Hejazi (b35) 2007; 33 Hua, Shuklab, Karniadakis, Kawaguchia (b12) 2024; 176 Lin, Chen (b28) 2023; 459 Batchelor (b2) 1967 David, Levi, Winternitz (b44) 1987; 76 Grimshaw (b39) 1979; 368 Wang, Cui (b22) 2022; 456 Wazwaz (b42) 2019; 182 Parker (b33) 1992; 438 Ko, Kuehl (b8) 1978; 40 Wu, Fang, Wang, Wu, Dai (b21) 2021; 152 Zhou, Pu, Chen (b25) 2023; 111 Tian, Niu, Li (b20) 2023; 111 Mo, Ling, Zeng (b23) 2022; 421 Shuning, Yong (b18) 2022; 457 Liu, Wang, Zhang, Lu, Liu (b24) 2023; 108 Fletcher (b36) 1983; 51 Pu, Chen (b16) 2024; 510 Grimshaw (b6) 1978; 86 Cohen-Tannoudji, Diu, Laloë (b3) 2019 Raissi, Perdikaris, Karniadakis (b10) 2019; 378 Dia, Chau, Lu, Zheng (b34) 2020; 43 David, Levi, Winternitz (b45) 1989; 80 Karniadakis, Kevrekidis, Lu (b13) 2021; 3 Yu, Lu, Meng, Karniadakis (b31) 2022; 393 Kadomtsev, Petviashvili (b43) 1970; Vol. 192 Keener, Sneyd (b4) 2009 Braun, Kivshar (b41) 2004 Lin, Chen (b17) 2023; 445 Fan (b40) 2002; 294 Awawdeh, Jaradat, Al-Shara (b38) 2012; 66 Li, Chen (b14) 2020; 72 Kakutani (b5) 1971; 30 Carslaw (b1) 1906 Wang, Yan (b19) 2021; 404 Hornik, Stinchcombe, White (b11) 1989; 2 Miao, Chen (b27) 2023; 456 He, Zhang, Ren, Sun (b30) 2016 Nishikawa, Kaw (b7) 1975; 50 Pu, Chen (b15) 2023; 454 Cole (b32) 1951; 9 Ali, Gardner, Gardner (b37) 1992; 100 Zhou (10.1016/j.physd.2025.134764_b26) 2024; 37 Raissi (10.1016/j.physd.2025.134764_b10) 2019; 378 Moghimi (10.1016/j.physd.2025.134764_b35) 2007; 33 Kakutani (10.1016/j.physd.2025.134764_b5) 1971; 30 Mo (10.1016/j.physd.2025.134764_b23) 2022; 421 Karniadakis (10.1016/j.physd.2025.134764_b13) 2021; 3 Hornik (10.1016/j.physd.2025.134764_b11) 1989; 2 Liu (10.1016/j.physd.2025.134764_b24) 2023; 108 Wang (10.1016/j.physd.2025.134764_b19) 2021; 404 Wang (10.1016/j.physd.2025.134764_b22) 2022; 456 Ott (10.1016/j.physd.2025.134764_b9) 1970; 13 Grimshaw (10.1016/j.physd.2025.134764_b6) 1978; 86 Parker (10.1016/j.physd.2025.134764_b33) 1992; 438 David (10.1016/j.physd.2025.134764_b44) 1987; 76 Ko (10.1016/j.physd.2025.134764_b8) 1978; 40 Tian (10.1016/j.physd.2025.134764_b20) 2023; 111 Wazwaz (10.1016/j.physd.2025.134764_b42) 2019; 182 Cohen-Tannoudji (10.1016/j.physd.2025.134764_b3) 2019 He (10.1016/j.physd.2025.134764_b30) 2016 Cole (10.1016/j.physd.2025.134764_b32) 1951; 9 Pu (10.1016/j.physd.2025.134764_b16) 2024; 510 Miao (10.1016/j.physd.2025.134764_b27) 2023; 456 Awawdeh (10.1016/j.physd.2025.134764_b38) 2012; 66 Zhou (10.1016/j.physd.2025.134764_b25) 2023; 111 Ali (10.1016/j.physd.2025.134764_b37) 1992; 100 Batchelor (10.1016/j.physd.2025.134764_b2) 1967 Wu (10.1016/j.physd.2025.134764_b21) 2021; 152 Shuning (10.1016/j.physd.2025.134764_b18) 2022; 457 Keener (10.1016/j.physd.2025.134764_b4) 2009 Kadomtsev (10.1016/j.physd.2025.134764_b43) 1970; Vol. 192 Pu (10.1016/j.physd.2025.134764_b15) 2023; 454 Hua (10.1016/j.physd.2025.134764_b12) 2024; 176 Yu (10.1016/j.physd.2025.134764_b31) 2022; 393 Fletcher (10.1016/j.physd.2025.134764_b36) 1983; 51 Nishikawa (10.1016/j.physd.2025.134764_b7) 1975; 50 Song (10.1016/j.physd.2025.134764_b29) 2023; 448 Lin (10.1016/j.physd.2025.134764_b17) 2023; 445 David (10.1016/j.physd.2025.134764_b45) 1989; 80 Güngör (10.1016/j.physd.2025.134764_b46) 2002; 276 Li (10.1016/j.physd.2025.134764_b14) 2020; 72 Grimshaw (10.1016/j.physd.2025.134764_b39) 1979; 368 Braun (10.1016/j.physd.2025.134764_b41) 2004 Fan (10.1016/j.physd.2025.134764_b40) 2002; 294 Lin (10.1016/j.physd.2025.134764_b28) 2023; 459 Carslaw (10.1016/j.physd.2025.134764_b1) 1906 Dia (10.1016/j.physd.2025.134764_b34) 2020; 43 |
| References_xml | – volume: 111 start-page: 16467 year: 2023 end-page: 16482 ident: b20 article-title: Mix-training physics-informed neural networks for high-order rogue waves of cmKdV equation publication-title: Nonlinear Dynam. – volume: 510 year: 2024 ident: b16 article-title: Lax pairs informed neural networks solving integrable systems publication-title: J. Comput. Phys. – volume: 2 start-page: 359 year: 1989 end-page: 366 ident: b11 article-title: Multilayer feedforward networks are universal approximators publication-title: Neural Netw. – volume: 456 year: 2023 ident: b27 article-title: VC-PINN: Variable coefficient physics-informed neural network for forward and inverse problems of PDEs with variable coefficient publication-title: Phys. D – volume: 182 start-page: 605 year: 2019 end-page: 610 ident: b42 article-title: The integrable time-dependent sine-Gordon equation with multiple optical kink solutions publication-title: Optik – volume: 9 start-page: 225 year: 1951 end-page: 236 ident: b32 article-title: On a quasi-linear parabolic equation occurring in aerodynamics publication-title: Quart. Appl. Math. – volume: 40 start-page: 233 year: 1978 end-page: 236 ident: b8 article-title: Korteweg–de Vries soliton in a slowly varying medium publication-title: Phys. Rev. Lett. – year: 1906 ident: b1 article-title: Introduction to the Mathematical Theory of the Conduction of Heat in Solids – year: 1967 ident: b2 article-title: An Introduction to Fluid Dynamics – volume: 176 year: 2024 ident: b12 article-title: Tackling the curse of dimensionality with physics-informed neural networks publication-title: Neural Netw. – volume: 448 year: 2023 ident: b29 article-title: Deep learning soliton dynamics and complex potentials recognition for 1D and 2D PT-symmetric saturable nonlinear Schrödinger equations publication-title: Phys. D – volume: 66 start-page: 40 year: 2012 ident: b38 article-title: Applications of a simplified bilinear method to ion-acoustic solitary waves in plasma publication-title: Eur. Phys. J. D – volume: 445 year: 2023 ident: b17 article-title: Physics-informed neural network methods based on Miura transformations and discovery of new localized wave solutions publication-title: Phys. D – volume: 76 start-page: 133 year: 1987 end-page: 168 ident: b44 article-title: Integrable nonlinear equations for water waves in straits of varying depth and width publication-title: Stud. Appl. Math. – volume: 51 start-page: 159 year: 1983 end-page: 188 ident: b36 article-title: A comparison of finite element and finite difference solutions of the one-and two-dimensional Burgers equations publication-title: J. Comput. Phys. – volume: 454 year: 2023 ident: b15 article-title: Complex dynamics on the one-dimensional quantum droplets via time piecewise PINNs publication-title: Phys. D – volume: 100 start-page: 325 year: 1992 end-page: 337 ident: b37 article-title: A collocation solution for Burgers’ equation using cubic B-spline finite elements publication-title: Comput. Methods Appl. Mech. Engrg. – volume: 276 start-page: 314 year: 2002 end-page: 328 ident: b46 article-title: Generalized Kadomtsev–Petviashvili equation with an infinite-dimensional symmetry algebra publication-title: J. Math. Anal. Appl. – volume: 33 start-page: 1756 year: 2007 end-page: 1761 ident: b35 article-title: Variational iteration method for solving generalized Burger–Fisher and Burger equations publication-title: Chaos – year: 2004 ident: b41 article-title: The Frenkel-Kontorova Model: Concepts, Methods, and Applications – volume: 457 year: 2022 ident: b18 article-title: A two-stage physics-informed neural network method based on conserved quantities and applications in localized wave solutions publication-title: J. Comput. Phys. – volume: 86 start-page: 415 year: 1978 end-page: 431 ident: b6 article-title: Long nonlinear internal waves in channels of arbitrary cross-section publication-title: J. Fluid Mech. – start-page: 630 year: 2016 end-page: 645 ident: b30 article-title: Identity Mappings in Deep Residual Networks – volume: 3 start-page: 422 year: 2021 end-page: 440 ident: b13 article-title: Physics-informed machine learning publication-title: Nat. Rev. Phys. – volume: 50 start-page: 455 year: 1975 end-page: 456 ident: b7 article-title: Propagation of solitary ion acoustic waves in inhomogeneous plasmas publication-title: Phys. Lett. A – volume: 459 year: 2023 ident: b28 article-title: Gradient-enhanced physics-informed neural networks based on transfer learning for inverse problems of the variable coefficient differential equations publication-title: Phys. D – volume: 72 year: 2020 ident: b14 article-title: Solving second-order nonlinear evolution partial differential equations using deep learning publication-title: Commun. Theor. Phys. (Beijing) – volume: 368 start-page: 359 year: 1979 end-page: 375 ident: b39 article-title: Slowly varying solitary waves. I. Korteweg–de Vries equation publication-title: Proc. R. Soc. Lond. Ser. A Math. Phys. Eng. Sci. – volume: Vol. 192 start-page: 753 year: 1970 end-page: 756 ident: b43 article-title: On the stability of solitary waves in weakly dispersing media publication-title: Doklady Akademii Nauk – volume: 80 start-page: 1 year: 1989 end-page: 23 ident: b45 article-title: Solitons in shallow seas of variable depth and in marine straits publication-title: Stud. Appl. Math. – volume: 421 year: 2022 ident: b23 article-title: Data-driven vector soliton solutions of coupled nonlinear Schrödinger equation using a deep learning algorithm publication-title: Phys. Lett. A – volume: 43 start-page: 2171 year: 2020 end-page: 2188 ident: b34 article-title: Mathematical studies of the solution of Burgers’ equations by adomian decomposition method publication-title: Math. Methods Appl. Sci. – volume: 30 start-page: 272 year: 1971 end-page: 276 ident: b5 article-title: Effect of an uneven bottom on gravity waves publication-title: J. Phys. Soc. Japan – volume: 438 start-page: 93 year: 1992 end-page: 108 ident: b33 article-title: On periodic solutions of the burgers equation: a unified approach publication-title: Proc. R. Soc. Lond. Ser. A Math. Phys. Eng. Sci. – year: 2019 ident: b3 article-title: Quantum Mechanics: Concepts and Applications – volume: 152 year: 2021 ident: b21 article-title: Predicting the dynamic process and model parameters of the vector optical solitons in birefringent fibers via the modified PINN publication-title: Chaos Solitons Fractals – year: 2009 ident: b4 article-title: Mathematical Physiology 1: Cellular Physiology – volume: 378 start-page: 686 year: 2019 end-page: 707 ident: b10 article-title: Physics-informed neural net- works: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations publication-title: J. Comput. Phys. – volume: 456 year: 2022 ident: b22 article-title: Prediction of the number of solitons for initial value of nonlinear Schrödinger equation based on the deep learning method publication-title: Phys. Lett. A – volume: 13 start-page: 1432 year: 1970 end-page: 1434 ident: b9 article-title: Damping of solitary waves publication-title: Phys. Fluids – volume: 404 year: 2021 ident: b19 article-title: Data-driven rogue waves and parameter discovery in the defocusing nonlinear Schrödinger equation with a potential using the PINN deeplearning publication-title: Phys. Lett. A – volume: 393 year: 2022 ident: b31 article-title: Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems publication-title: Comput. Methods Appl. Mech. Engrg. – volume: 294 start-page: 26 year: 2002 end-page: 30 ident: b40 article-title: Auto-Bäcklund transformation and similarity reductions for general variable coefficient KdV equations publication-title: Phys. Lett. A – volume: 111 start-page: 14667 year: 2023 end-page: 14693 ident: b25 article-title: Data-driven forward–inverse problems for the variable coefficients Hirota equation using deep learning method publication-title: Nonlinear Dynam. – volume: 37 start-page: 511 year: 2024 end-page: 544 ident: b26 article-title: Parallel physics-informed neural networks method with regularization strategies for the forward-inverse problems of the variable coefficient modified KdV equation publication-title: J. Syst. Sci. Complex. – volume: 108 year: 2023 ident: b24 article-title: Prediction of phase transition and time-varying dynamics of the (2+1)-dimensional Boussinesq equation by the parameter-integrated physics-informed neural networks with phase domain decomposition publication-title: Phys. Rev. E – volume: 421 year: 2022 ident: 10.1016/j.physd.2025.134764_b23 article-title: Data-driven vector soliton solutions of coupled nonlinear Schrödinger equation using a deep learning algorithm publication-title: Phys. Lett. A doi: 10.1016/j.physleta.2021.127739 – volume: 456 year: 2022 ident: 10.1016/j.physd.2025.134764_b22 article-title: Prediction of the number of solitons for initial value of nonlinear Schrödinger equation based on the deep learning method publication-title: Phys. Lett. A doi: 10.1016/j.physleta.2022.128536 – volume: 448 year: 2023 ident: 10.1016/j.physd.2025.134764_b29 article-title: Deep learning soliton dynamics and complex potentials recognition for 1D and 2D PT-symmetric saturable nonlinear Schrödinger equations publication-title: Phys. D doi: 10.1016/j.physd.2023.133729 – volume: 30 start-page: 272 year: 1971 ident: 10.1016/j.physd.2025.134764_b5 article-title: Effect of an uneven bottom on gravity waves publication-title: J. Phys. Soc. Japan doi: 10.1143/JPSJ.30.272 – volume: 108 year: 2023 ident: 10.1016/j.physd.2025.134764_b24 article-title: Prediction of phase transition and time-varying dynamics of the (2+1)-dimensional Boussinesq equation by the parameter-integrated physics-informed neural networks with phase domain decomposition publication-title: Phys. Rev. E doi: 10.1103/PhysRevE.108.045303 – year: 2004 ident: 10.1016/j.physd.2025.134764_b41 – volume: 76 start-page: 133 issue: 2 year: 1987 ident: 10.1016/j.physd.2025.134764_b44 article-title: Integrable nonlinear equations for water waves in straits of varying depth and width publication-title: Stud. Appl. Math. doi: 10.1002/sapm1987762133 – volume: 459 year: 2023 ident: 10.1016/j.physd.2025.134764_b28 article-title: Gradient-enhanced physics-informed neural networks based on transfer learning for inverse problems of the variable coefficient differential equations publication-title: Phys. D – volume: 176 year: 2024 ident: 10.1016/j.physd.2025.134764_b12 article-title: Tackling the curse of dimensionality with physics-informed neural networks publication-title: Neural Netw. – volume: 51 start-page: 159 issue: 1 year: 1983 ident: 10.1016/j.physd.2025.134764_b36 article-title: A comparison of finite element and finite difference solutions of the one-and two-dimensional Burgers equations publication-title: J. Comput. Phys. doi: 10.1016/0021-9991(83)90085-2 – volume: 152 year: 2021 ident: 10.1016/j.physd.2025.134764_b21 article-title: Predicting the dynamic process and model parameters of the vector optical solitons in birefringent fibers via the modified PINN publication-title: Chaos Solitons Fractals doi: 10.1016/j.chaos.2021.111393 – volume: 445 year: 2023 ident: 10.1016/j.physd.2025.134764_b17 article-title: Physics-informed neural network methods based on Miura transformations and discovery of new localized wave solutions publication-title: Phys. D doi: 10.1016/j.physd.2022.133629 – volume: 2 start-page: 359 year: 1989 ident: 10.1016/j.physd.2025.134764_b11 article-title: Multilayer feedforward networks are universal approximators publication-title: Neural Netw. doi: 10.1016/0893-6080(89)90020-8 – volume: 457 year: 2022 ident: 10.1016/j.physd.2025.134764_b18 article-title: A two-stage physics-informed neural network method based on conserved quantities and applications in localized wave solutions publication-title: J. Comput. Phys. – volume: 393 year: 2022 ident: 10.1016/j.physd.2025.134764_b31 article-title: Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems publication-title: Comput. Methods Appl. Mech. Engrg. doi: 10.1016/j.cma.2022.114823 – volume: 182 start-page: 605 year: 2019 ident: 10.1016/j.physd.2025.134764_b42 article-title: The integrable time-dependent sine-Gordon equation with multiple optical kink solutions publication-title: Optik doi: 10.1016/j.ijleo.2019.01.018 – volume: 33 start-page: 1756 issue: 5 year: 2007 ident: 10.1016/j.physd.2025.134764_b35 article-title: Variational iteration method for solving generalized Burger–Fisher and Burger equations publication-title: Chaos – volume: 111 start-page: 14667 year: 2023 ident: 10.1016/j.physd.2025.134764_b25 article-title: Data-driven forward–inverse problems for the variable coefficients Hirota equation using deep learning method publication-title: Nonlinear Dynam. doi: 10.1007/s11071-023-08641-1 – volume: 43 start-page: 2171 issue: 5 year: 2020 ident: 10.1016/j.physd.2025.134764_b34 article-title: Mathematical studies of the solution of Burgers’ equations by adomian decomposition method publication-title: Math. Methods Appl. Sci. doi: 10.1002/mma.5982 – volume: 86 start-page: 415 year: 1978 ident: 10.1016/j.physd.2025.134764_b6 article-title: Long nonlinear internal waves in channels of arbitrary cross-section publication-title: J. Fluid Mech. doi: 10.1017/S0022112078001214 – volume: 454 year: 2023 ident: 10.1016/j.physd.2025.134764_b15 article-title: Complex dynamics on the one-dimensional quantum droplets via time piecewise PINNs publication-title: Phys. D doi: 10.1016/j.physd.2023.133851 – volume: 510 year: 2024 ident: 10.1016/j.physd.2025.134764_b16 article-title: Lax pairs informed neural networks solving integrable systems publication-title: J. Comput. Phys. doi: 10.1016/j.jcp.2024.113090 – volume: 438 start-page: 93 year: 1992 ident: 10.1016/j.physd.2025.134764_b33 article-title: On periodic solutions of the burgers equation: a unified approach publication-title: Proc. R. Soc. Lond. Ser. A Math. Phys. Eng. Sci. – volume: 378 start-page: 686 year: 2019 ident: 10.1016/j.physd.2025.134764_b10 article-title: Physics-informed neural net- works: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations publication-title: J. Comput. Phys. doi: 10.1016/j.jcp.2018.10.045 – volume: 80 start-page: 1 issue: 1 year: 1989 ident: 10.1016/j.physd.2025.134764_b45 article-title: Solitons in shallow seas of variable depth and in marine straits publication-title: Stud. Appl. Math. doi: 10.1002/sapm19898011 – volume: 456 year: 2023 ident: 10.1016/j.physd.2025.134764_b27 article-title: VC-PINN: Variable coefficient physics-informed neural network for forward and inverse problems of PDEs with variable coefficient publication-title: Phys. D doi: 10.1016/j.physd.2023.133945 – start-page: 630 year: 2016 ident: 10.1016/j.physd.2025.134764_b30 – volume: 72 year: 2020 ident: 10.1016/j.physd.2025.134764_b14 article-title: Solving second-order nonlinear evolution partial differential equations using deep learning publication-title: Commun. Theor. Phys. (Beijing) – volume: 66 start-page: 40 year: 2012 ident: 10.1016/j.physd.2025.134764_b38 article-title: Applications of a simplified bilinear method to ion-acoustic solitary waves in plasma publication-title: Eur. Phys. J. D doi: 10.1140/epjd/e2011-20518-0 – volume: 294 start-page: 26 issue: 1 year: 2002 ident: 10.1016/j.physd.2025.134764_b40 article-title: Auto-Bäcklund transformation and similarity reductions for general variable coefficient KdV equations publication-title: Phys. Lett. A doi: 10.1016/S0375-9601(02)00033-6 – volume: Vol. 192 start-page: 753 year: 1970 ident: 10.1016/j.physd.2025.134764_b43 article-title: On the stability of solitary waves in weakly dispersing media – year: 1967 ident: 10.1016/j.physd.2025.134764_b2 – volume: 40 start-page: 233 year: 1978 ident: 10.1016/j.physd.2025.134764_b8 article-title: Korteweg–de Vries soliton in a slowly varying medium publication-title: Phys. Rev. Lett. doi: 10.1103/PhysRevLett.40.233 – volume: 404 year: 2021 ident: 10.1016/j.physd.2025.134764_b19 article-title: Data-driven rogue waves and parameter discovery in the defocusing nonlinear Schrödinger equation with a potential using the PINN deeplearning publication-title: Phys. Lett. A doi: 10.1016/j.physleta.2021.127408 – volume: 111 start-page: 16467 year: 2023 ident: 10.1016/j.physd.2025.134764_b20 article-title: Mix-training physics-informed neural networks for high-order rogue waves of cmKdV equation publication-title: Nonlinear Dynam. doi: 10.1007/s11071-023-08712-3 – volume: 37 start-page: 511 year: 2024 ident: 10.1016/j.physd.2025.134764_b26 article-title: Parallel physics-informed neural networks method with regularization strategies for the forward-inverse problems of the variable coefficient modified KdV equation publication-title: J. Syst. Sci. Complex. doi: 10.1007/s11424-024-3467-7 – volume: 368 start-page: 359 year: 1979 ident: 10.1016/j.physd.2025.134764_b39 article-title: Slowly varying solitary waves. I. Korteweg–de Vries equation publication-title: Proc. R. Soc. Lond. Ser. A Math. Phys. Eng. Sci. – volume: 276 start-page: 314 issue: 1 year: 2002 ident: 10.1016/j.physd.2025.134764_b46 article-title: Generalized Kadomtsev–Petviashvili equation with an infinite-dimensional symmetry algebra publication-title: J. Math. Anal. Appl. doi: 10.1016/S0022-247X(02)00445-6 – year: 2019 ident: 10.1016/j.physd.2025.134764_b3 – volume: 100 start-page: 325 issue: 3 year: 1992 ident: 10.1016/j.physd.2025.134764_b37 article-title: A collocation solution for Burgers’ equation using cubic B-spline finite elements publication-title: Comput. Methods Appl. Mech. Engrg. doi: 10.1016/0045-7825(92)90088-2 – volume: 13 start-page: 1432 year: 1970 ident: 10.1016/j.physd.2025.134764_b9 article-title: Damping of solitary waves publication-title: Phys. Fluids doi: 10.1063/1.1693097 – volume: 3 start-page: 422 year: 2021 ident: 10.1016/j.physd.2025.134764_b13 article-title: Physics-informed machine learning publication-title: Nat. Rev. Phys. doi: 10.1038/s42254-021-00314-5 – volume: 9 start-page: 225 issue: 3 year: 1951 ident: 10.1016/j.physd.2025.134764_b32 article-title: On a quasi-linear parabolic equation occurring in aerodynamics publication-title: Quart. Appl. Math. doi: 10.1090/qam/42889 – year: 1906 ident: 10.1016/j.physd.2025.134764_b1 – year: 2009 ident: 10.1016/j.physd.2025.134764_b4 – volume: 50 start-page: 455 year: 1975 ident: 10.1016/j.physd.2025.134764_b7 article-title: Propagation of solitary ion acoustic waves in inhomogeneous plasmas publication-title: Phys. Lett. A doi: 10.1016/0375-9601(75)90124-3 |
| SSID | ssj0001737 |
| Score | 2.4729505 |
| Snippet | Physics-informed neural network (PINN) is a powerful emerging method for studying forward-inverse problems of partial differential equations (PDEs), even from... |
| SourceID | crossref elsevier |
| SourceType | Index Database Publisher |
| StartPage | 134764 |
| SubjectTerms | Data-driven solution Function discovery Physics-informed neural network Residual network Variable coefficient equation |
| Title | Gradient-enhanced PINN with residual unit for studying forward-inverse problems of variable coefficient equations |
| URI | https://dx.doi.org/10.1016/j.physd.2025.134764 |
| Volume | 481 |
| WOSCitedRecordID | wos001516751600001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| journalDatabaseRights | – providerCode: PRVESC databaseName: Elsevier SD Freedom Collection Journals 2021 issn: 0167-2789 databaseCode: AIEXJ dateStart: 19950101 customDbUrl: isFulltext: true dateEnd: 99991231 titleUrlDefault: https://www.sciencedirect.com omitProxy: false ssIdentifier: ssj0001737 providerName: Elsevier |
| link | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1LT9tAEF7xKBI9IKBF0ALaA8c68mu93iNClIdQxAFQ2otl70MkBxuSGPHzmX0lhlSoHLhYlmWvLX-fZ2fG38widFSSMleUAQJCJkHKkyhgSVgGLK1EVoJ1pOZH-90V7ffzwYBdu4YKE7OcAK3r_PmZPXwq1HAMwNalsx-AezYoHIB9AB22ADts_wv4s7FRcU0DWd_b3_sQufdtwhVia1t81cKXbBSGpr2sU1MaAe2w1kINXT5lVpoxSo8nCKhNiRVvpGk5oQUE8rHtZPucf3ttYe_NhcR_75vWzG_tMLhs51w8cWUhfxo3d7rUQ0xcDV4nGwlWVpfSds1papdgcQZRV6raNuULttqmDUY9ncLRPVtj0puf_boz9psZa6Yj9BK1UWEGKfQghR1kGa3GlDCw1avHF6eDy9n0HFHbSNU_u29FZUR_C8_yb3el44LcbKINFzvgY4v5FlqS9Tb62ukouY3WLACTb-hxgQdY8wBrHmDPA6x5gAF67HmA3_AAex7gRmHPA9zhAZ7x4Du6_X16c3IeuPU1Ag6B5DSIM1Vy8FwE-CQSHDURVpEImYhVLCHsJDxUKStVllcR4xB404iVpqIIzs14QpMdtFI3tdxFWMRSVJXgQoksJTmv4lLRSIVESZLxnOyhX_41Fg-2jUrxDnh7KPOvunCeoPXwCiDPexf--Nh9fqL1Oa_30cp03MoD9IU_TYeT8aFjzgsWA4MM |
| linkProvider | Elsevier |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Gradient-enhanced+PINN+with+residual+unit+for+studying+forward-inverse+problems+of+variable+coefficient+equations&rft.jtitle=Physica.+D&rft.au=Zhou%2C+Hui-Juan&rft.au=Chen%2C+Yong&rft.date=2025-11-01&rft.issn=0167-2789&rft.volume=481&rft.spage=134764&rft_id=info:doi/10.1016%2Fj.physd.2025.134764&rft.externalDBID=n%2Fa&rft.externalDocID=10_1016_j_physd_2025_134764 |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0167-2789&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0167-2789&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0167-2789&client=summon |