A modified iterative PINN algorithm for strongly coupled system of boundary layer originated convection diffusion reaction problems in MHD flows with analysis

In this work, we design an iteration-based unsupervised neural network algorithm and the theoretical bound of the loss function through Barron space to solve reaction–convection–diffusion-based magnetohydrodynamic (MHD) coupled systems for 1D and 2D problems. Our algorithm is also be able to capture...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Networks and heterogeneous media Jg. 20; H. 3; S. 1026 - 1060
Hauptverfasser: Patawari, Arihant, Kumar, Shridhar, Das, Pratibhamoy
Format: Journal Article
Sprache:Englisch
Veröffentlicht: AIMS Press 01.01.2025
Schlagworte:
ISSN:1556-1801
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Abstract In this work, we design an iteration-based unsupervised neural network algorithm and the theoretical bound of the loss function through Barron space to solve reaction–convection–diffusion-based magnetohydrodynamic (MHD) coupled systems for 1D and 2D problems. Our algorithm is also be able to capture the presence of boundary layers. In general, these systems are characterized by strong coupling in the reaction and convection processes and involve non-diagonally dominant matrices in the context of convection coupling. Traditional numerical techniques face challenges in approximating these problems due to the failure of the required maximum principle, which is used for the well-posedness and further convergence analysis of numerical solutions. We specifically provide a new modified iterative physics-informed neural network (MI-PINN)-based unsupervised deep learning algorithm to capture the layer behavior of singularly perturbed strongly coupled steady-state problems, appearing in MHD flows where theoretical analysis and numerical methods are limited. A different analysis based on the sigmoid activation function is provided for the steady-state case, which shows that the empirical loss under the $ L^2 $ norm is bounded and converges for the two layer-based networks- whenever the solution lies in the Barron space. Additionally, the proposed algorithm improves the neural network's output without using the boundary layer functions a priori and does not use hard constraints or interpolation with the neural network's solution. The experimental results show that the proposed algorithm performs very well for MHD flows appearing in the form of strongly coupled systems.
AbstractList In this work, we design an iteration-based unsupervised neural network algorithm and the theoretical bound of the loss function through Barron space to solve reaction–convection–diffusion-based magnetohydrodynamic (MHD) coupled systems for 1D and 2D problems. Our algorithm is also be able to capture the presence of boundary layers. In general, these systems are characterized by strong coupling in the reaction and convection processes and involve non-diagonally dominant matrices in the context of convection coupling. Traditional numerical techniques face challenges in approximating these problems due to the failure of the required maximum principle, which is used for the well-posedness and further convergence analysis of numerical solutions. We specifically provide a new modified iterative physics-informed neural network (MI-PINN)-based unsupervised deep learning algorithm to capture the layer behavior of singularly perturbed strongly coupled steady-state problems, appearing in MHD flows where theoretical analysis and numerical methods are limited. A different analysis based on the sigmoid activation function is provided for the steady-state case, which shows that the empirical loss under the $ L^2 $ norm is bounded and converges for the two layer-based networks- whenever the solution lies in the Barron space. Additionally, the proposed algorithm improves the neural network's output without using the boundary layer functions a priori and does not use hard constraints or interpolation with the neural network's solution. The experimental results show that the proposed algorithm performs very well for MHD flows appearing in the form of strongly coupled systems.
Author Kumar, Shridhar
Patawari, Arihant
Das, Pratibhamoy
Author_xml – sequence: 1
  givenname: Arihant
  surname: Patawari
  fullname: Patawari, Arihant
– sequence: 2
  givenname: Shridhar
  surname: Kumar
  fullname: Kumar, Shridhar
– sequence: 3
  givenname: Pratibhamoy
  surname: Das
  fullname: Das, Pratibhamoy
BookMark eNo9UctOJCEUZeEkajsrf4C9aQeKKrpYGp-dqDOLcV25BZcWQ0EHaE39jN8qPW1mdW9OziMn55QchRiQkHPOLoUS7a_wOl02rOlY2x2RE951csl7xo_Jac5vjLVixcQJ-byiUzTOOjTUFUxQ3DvSP-vnZwp-E5MrrxO1MdFcUgwbP1Mdd1tf2XnOBScaLR3jLhhIM_UwY6JVtHEBSuXoGN5RFxcDrRl2l_dfQjhA2xRHj1OmLtCnhxtqffzI9KNGUgjg5-zyGflhwWf8-X0X5OXu9u_1w_Lx9_36-upxqRslyxKV7nknBHaSg8TGIsdaEGDVtr01WiqtpDBGgFJgNDDVKbNC2WODUksQC7I--JoIb8M2uan2GSK44R8Q02aAVJz2OCA2oEeFthdNa5lQvFdGjkaMKyN4jVmQi4OXTjHnhPa_H2fDfpqhTjN8TyO-ADpFimg
Cites_doi 10.1007/BF02551274
10.1016/S0168-9274(97)00106-2
10.1016/j.apnum.2024.09.020
10.1016/j.jcp.2010.08.034
10.1002/zamm.201000153
10.1016/j.jcp.2018.10.045
10.1016/S0377-0427(00)00260-0
10.1016/j.cma.2022.115616
10.1093/imanum/drac085
10.1016/j.cam.2024.116223
10.1016/j.apnum.2019.07.003
10.1016/j.apnum.2023.10.003
10.1090/gsm/075
10.1016/j.jfranklin.2019.12.013
10.4208/cicp.OA-2020-0193
10.1007/s10444-019-09669-x
10.1016/j.apnum.2023.01.003
10.1007/s11071-015-1894-7
10.1016/j.jcp.2023.112702
10.1007/s00607-006-0215-x
10.1002/fld.5294
10.1007/s10546-021-00666-6
10.1007/s10915-022-01939-z
10.1016/j.jcp.2022.111768
10.1093/imanum/drab093
10.1016/j.jcp.2019.109136
10.1016/j.cnsns.2025.108627
10.1080/02331934.2025.2534119
10.1016/j.cam.2017.11.026
10.1109/18.256500
10.1016/S0096-3003(01)00291-0
10.1016/j.apnum.2022.09.012
10.1109/TSMC.2022.3152505
10.4208/ijnam2025-1020
10.1080/00036811.2024.2302405
10.1111/sapm.12763
10.2478/cmam-2009-0010
10.1016/j.bulsci.2025.103637
10.1007/s00365-021-09549-y
10.1007/s10483-024-3149-8
10.1016/j.cma.2020.113547
10.1007/s10444-007-9058-z
10.3934/nhm.2024048
10.1016/j.apnum.2019.08.028
10.1145/3689037
10.1016/j.aml.2012.02.018
10.1051/proc/202373048
ContentType Journal Article
CorporateAuthor Department of Mathematics, Indian Institute of Technology Patna, India
School of Mathematics, Indian Institute of Science Education and Research Thiruvananthapuram, India
CorporateAuthor_xml – name: School of Mathematics, Indian Institute of Science Education and Research Thiruvananthapuram, India
– name: Department of Mathematics, Indian Institute of Technology Patna, India
DBID AAYXX
CITATION
DOA
DOI 10.3934/nhm.2025045
DatabaseName CrossRef
Directory of Open Access Journals (DOAJ)
DatabaseTitle CrossRef
DatabaseTitleList
Database_xml – sequence: 1
  dbid: DOA
  name: DOAJ Directory of Open Access Journals
  url: https://www.doaj.org/
  sourceTypes: Open Website
DeliveryMethod fulltext_linktorsrc
Discipline Applied Sciences
EndPage 1060
ExternalDocumentID oai_doaj_org_article_ee2acb9ef8324f039189d6bd3b7d3196
10_3934_nhm_2025045
GroupedDBID .4S
.DC
123
AAYXX
AENEX
ALMA_UNASSIGNED_HOLDINGS
AMVHM
ARCSS
CITATION
DU5
E3Z
EBS
EJD
GROUPED_DOAJ
J9A
RAN
TUS
ID FETCH-LOGICAL-c296t-e9c81533e561a6e2fe1e437aa7448fdc69c963dd3a99adca0959d7e68e2e6c6a3
IEDL.DBID DOA
ISICitedReferencesCount 2
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001587183400004&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 1556-1801
IngestDate Mon Nov 03 22:05:41 EST 2025
Sat Nov 29 07:26:43 EST 2025
IsDoiOpenAccess true
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Issue 3
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c296t-e9c81533e561a6e2fe1e437aa7448fdc69c963dd3a99adca0959d7e68e2e6c6a3
OpenAccessLink https://doaj.org/article/ee2acb9ef8324f039189d6bd3b7d3196
PageCount 35
ParticipantIDs doaj_primary_oai_doaj_org_article_ee2acb9ef8324f039189d6bd3b7d3196
crossref_primary_10_3934_nhm_2025045
PublicationCentury 2000
PublicationDate 2025-01-01
PublicationDateYYYYMMDD 2025-01-01
PublicationDate_xml – month: 01
  year: 2025
  text: 2025-01-01
  day: 01
PublicationDecade 2020
PublicationTitle Networks and heterogeneous media
PublicationYear 2025
Publisher AIMS Press
Publisher_xml – name: AIMS Press
References key-10.3934/nhm.2025045-9
key-10.3934/nhm.2025045-7
key-10.3934/nhm.2025045-8
key-10.3934/nhm.2025045-5
key-10.3934/nhm.2025045-41
key-10.3934/nhm.2025045-6
key-10.3934/nhm.2025045-42
key-10.3934/nhm.2025045-3
key-10.3934/nhm.2025045-43
key-10.3934/nhm.2025045-4
key-10.3934/nhm.2025045-44
key-10.3934/nhm.2025045-1
key-10.3934/nhm.2025045-2
key-10.3934/nhm.2025045-40
key-10.3934/nhm.2025045-38
key-10.3934/nhm.2025045-39
key-10.3934/nhm.2025045-34
key-10.3934/nhm.2025045-35
key-10.3934/nhm.2025045-36
key-10.3934/nhm.2025045-37
key-10.3934/nhm.2025045-52
key-10.3934/nhm.2025045-53
key-10.3934/nhm.2025045-10
key-10.3934/nhm.2025045-54
key-10.3934/nhm.2025045-11
key-10.3934/nhm.2025045-55
key-10.3934/nhm.2025045-50
key-10.3934/nhm.2025045-51
key-10.3934/nhm.2025045-49
key-10.3934/nhm.2025045-45
key-10.3934/nhm.2025045-46
key-10.3934/nhm.2025045-47
key-10.3934/nhm.2025045-48
key-10.3934/nhm.2025045-63
key-10.3934/nhm.2025045-20
key-10.3934/nhm.2025045-64
key-10.3934/nhm.2025045-21
key-10.3934/nhm.2025045-65
key-10.3934/nhm.2025045-22
key-10.3934/nhm.2025045-66
key-10.3934/nhm.2025045-60
key-10.3934/nhm.2025045-61
key-10.3934/nhm.2025045-62
key-10.3934/nhm.2025045-16
key-10.3934/nhm.2025045-17
key-10.3934/nhm.2025045-18
key-10.3934/nhm.2025045-19
key-10.3934/nhm.2025045-12
key-10.3934/nhm.2025045-56
key-10.3934/nhm.2025045-13
key-10.3934/nhm.2025045-57
key-10.3934/nhm.2025045-14
key-10.3934/nhm.2025045-58
key-10.3934/nhm.2025045-15
key-10.3934/nhm.2025045-59
key-10.3934/nhm.2025045-30
key-10.3934/nhm.2025045-31
key-10.3934/nhm.2025045-32
key-10.3934/nhm.2025045-33
key-10.3934/nhm.2025045-27
key-10.3934/nhm.2025045-28
key-10.3934/nhm.2025045-29
key-10.3934/nhm.2025045-23
key-10.3934/nhm.2025045-67
key-10.3934/nhm.2025045-24
key-10.3934/nhm.2025045-25
key-10.3934/nhm.2025045-26
References_xml – ident: key-10.3934/nhm.2025045-50
– ident: key-10.3934/nhm.2025045-48
– ident: key-10.3934/nhm.2025045-35
  doi: 10.1007/BF02551274
– ident: key-10.3934/nhm.2025045-5
– ident: key-10.3934/nhm.2025045-21
– ident: key-10.3934/nhm.2025045-31
– ident: key-10.3934/nhm.2025045-20
  doi: 10.1016/S0168-9274(97)00106-2
– ident: key-10.3934/nhm.2025045-29
  doi: 10.1016/j.apnum.2024.09.020
– ident: key-10.3934/nhm.2025045-63
– ident: key-10.3934/nhm.2025045-16
– ident: key-10.3934/nhm.2025045-8
  doi: 10.1016/j.jcp.2010.08.034
– ident: key-10.3934/nhm.2025045-64
  doi: 10.1002/zamm.201000153
– ident: key-10.3934/nhm.2025045-24
  doi: 10.1016/j.jcp.2018.10.045
– ident: key-10.3934/nhm.2025045-58
  doi: 10.1016/S0377-0427(00)00260-0
– ident: key-10.3934/nhm.2025045-26
  doi: 10.1016/j.cma.2022.115616
– ident: key-10.3934/nhm.2025045-38
  doi: 10.1093/imanum/drac085
– ident: key-10.3934/nhm.2025045-28
– ident: key-10.3934/nhm.2025045-40
  doi: 10.1016/j.cam.2024.116223
– ident: key-10.3934/nhm.2025045-6
– ident: key-10.3934/nhm.2025045-30
– ident: key-10.3934/nhm.2025045-2
  doi: 10.1016/j.apnum.2019.07.003
– ident: key-10.3934/nhm.2025045-12
  doi: 10.1016/j.apnum.2023.10.003
– ident: key-10.3934/nhm.2025045-46
  doi: 10.1090/gsm/075
– ident: key-10.3934/nhm.2025045-1
  doi: 10.1016/j.jfranklin.2019.12.013
– ident: key-10.3934/nhm.2025045-41
– ident: key-10.3934/nhm.2025045-36
  doi: 10.4208/cicp.OA-2020-0193
– ident: key-10.3934/nhm.2025045-13
  doi: 10.1007/s10444-019-09669-x
– ident: key-10.3934/nhm.2025045-53
  doi: 10.1016/j.apnum.2023.01.003
– ident: key-10.3934/nhm.2025045-55
– ident: key-10.3934/nhm.2025045-59
– ident: key-10.3934/nhm.2025045-18
  doi: 10.1007/s11071-015-1894-7
– ident: key-10.3934/nhm.2025045-27
  doi: 10.1016/j.jcp.2023.112702
– ident: key-10.3934/nhm.2025045-56
  doi: 10.1007/s00607-006-0215-x
– ident: key-10.3934/nhm.2025045-7
– ident: key-10.3934/nhm.2025045-22
  doi: 10.1002/fld.5294
– ident: key-10.3934/nhm.2025045-14
  doi: 10.1007/s10546-021-00666-6
– ident: key-10.3934/nhm.2025045-65
– ident: key-10.3934/nhm.2025045-32
  doi: 10.1007/s10915-022-01939-z
– ident: key-10.3934/nhm.2025045-25
  doi: 10.1016/j.jcp.2022.111768
– ident: key-10.3934/nhm.2025045-37
  doi: 10.1093/imanum/drab093
– ident: key-10.3934/nhm.2025045-10
– ident: key-10.3934/nhm.2025045-39
  doi: 10.1016/j.jcp.2019.109136
– ident: key-10.3934/nhm.2025045-42
– ident: key-10.3934/nhm.2025045-51
  doi: 10.1016/j.cnsns.2025.108627
– ident: key-10.3934/nhm.2025045-61
  doi: 10.1080/02331934.2025.2534119
– ident: key-10.3934/nhm.2025045-60
  doi: 10.1016/j.cam.2017.11.026
– ident: key-10.3934/nhm.2025045-4
– ident: key-10.3934/nhm.2025045-66
  doi: 10.1109/18.256500
– ident: key-10.3934/nhm.2025045-11
  doi: 10.1016/S0096-3003(01)00291-0
– ident: key-10.3934/nhm.2025045-52
  doi: 10.1016/j.apnum.2022.09.012
– ident: key-10.3934/nhm.2025045-3
  doi: 10.1109/TSMC.2022.3152505
– ident: key-10.3934/nhm.2025045-47
– ident: key-10.3934/nhm.2025045-19
  doi: 10.4208/ijnam2025-1020
– ident: key-10.3934/nhm.2025045-43
  doi: 10.1080/00036811.2024.2302405
– ident: key-10.3934/nhm.2025045-23
  doi: 10.1111/sapm.12763
– ident: key-10.3934/nhm.2025045-57
  doi: 10.2478/cmam-2009-0010
– ident: key-10.3934/nhm.2025045-17
  doi: 10.1016/j.bulsci.2025.103637
– ident: key-10.3934/nhm.2025045-49
  doi: 10.1007/s00365-021-09549-y
– ident: key-10.3934/nhm.2025045-45
  doi: 10.1007/s10483-024-3149-8
– ident: key-10.3934/nhm.2025045-44
  doi: 10.1016/j.cma.2020.113547
– ident: key-10.3934/nhm.2025045-54
  doi: 10.1007/s10444-007-9058-z
– ident: key-10.3934/nhm.2025045-15
– ident: key-10.3934/nhm.2025045-67
  doi: 10.3934/nhm.2024048
– ident: key-10.3934/nhm.2025045-9
  doi: 10.1016/j.apnum.2019.08.028
– ident: key-10.3934/nhm.2025045-34
  doi: 10.1145/3689037
– ident: key-10.3934/nhm.2025045-62
  doi: 10.1016/j.aml.2012.02.018
– ident: key-10.3934/nhm.2025045-33
  doi: 10.1051/proc/202373048
SSID ssj0043703
Score 2.341254
Snippet In this work, we design an iteration-based unsupervised neural network algorithm and the theoretical bound of the loss function through Barron space to solve...
SourceID doaj
crossref
SourceType Open Website
Index Database
StartPage 1026
SubjectTerms barron space
boundary layer in singular perturbation
convergence analysis
mi-pinn
multiple scale problems
multiple scale problems in 1d and 2d
neural network
pinn
strongly coupled system
unsteady mhd flow
Title A modified iterative PINN algorithm for strongly coupled system of boundary layer originated convection diffusion reaction problems in MHD flows with analysis
URI https://doaj.org/article/ee2acb9ef8324f039189d6bd3b7d3196
Volume 20
WOSCitedRecordID wos001587183400004&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: PRVAON
  databaseName: DOAJ Directory of Open Access Journals
  issn: 1556-1801
  databaseCode: DOA
  dateStart: 20230101
  customDbUrl:
  isFulltext: true
  dateEnd: 99991231
  titleUrlDefault: https://www.doaj.org/
  omitProxy: false
  ssIdentifier: ssj0043703
  providerName: Directory of Open Access Journals
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrV07bxsxDBaKoEOX9I0mfYBD10N8j-ik0X0EGVqjQwt4O-goyjZwuQt8dgr_mf7Wkie1cKcsXQVBEMRP1EdC_KjUeyxnLfo2z4zDNqu0NpmzwWSe0UUFVsyCp0LhL_ViYZZL--2o1Zf8CYvywPHgLogKXsVSYOhVQfTMjfW69WVbe4GPeN9Zbf8EU9EHV2U99UTmx1JnOTvhWJlX2rK66NdSgC7KXZf_vEVHkv3T23L1RJ0mUgjzuJmn6gH1z9TjRBAhXb_xufo1h5vBb4KMRjlk9lXAgfkCXLcaOM5f3wCzUBglwb3qDoDD_rbj2VGwGYYA7dRHaXuAzjHbhtQYi3knTB_QpzIHkLYpe8mjAXPKOJQ6z4yw6eHr9ScI3fBzBEnjgkvCJi_Uj6vP3z9eZ6nBQoaF1buMLBrhe8QkymkqAuXER-dczUFb8Kgt8v30vnTWOo9Ocoa-Jm2oII3alS_VST_09EqBxdZhjnWZt1SZS22wmNmZZ_aAwZAJZwyLdNTNbdTRaDj-EIs0bJEmWeRMfRAz_J0i4tfTAEOiSZBo7oPE-f9Y5LV6JHuK2ZY36mS33dNb9RDvdptx-25C22-jNd60
linkProvider Directory of Open Access Journals
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=A+modified+iterative+PINN+algorithm+for+strongly+coupled+system+of+boundary+layer+originated+convection+diffusion+reaction+problems+in+MHD+flows+with+analysis&rft.jtitle=Networks+and+heterogeneous+media&rft.au=Arihant+Patawari&rft.au=Shridhar+Kumar&rft.au=Pratibhamoy+Das&rft.date=2025-01-01&rft.pub=AIMS+Press&rft.issn=1556-1801&rft.volume=20&rft.issue=3&rft.spage=1026&rft.epage=1060&rft_id=info:doi/10.3934%2Fnhm.2025045&rft.externalDBID=DOA&rft.externalDocID=oai_doaj_org_article_ee2acb9ef8324f039189d6bd3b7d3196
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1556-1801&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1556-1801&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1556-1801&client=summon