<?xml version="1.0" encoding="utf-8"?>
<TEI xmlns="http://www.tei-c.org/ns/1.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:hal="http://hal.archives-ouvertes.fr/" xmlns:gml="http://www.opengis.net/gml/3.3/" xmlns:gmlce="http://www.opengis.net/gml/3.3/ce" version="1.1" xsi:schemaLocation="http://www.tei-c.org/ns/1.0 http://api.archives-ouvertes.fr/documents/aofr-sword.xsd">
  <teiHeader>
    <fileDesc>
      <titleStmt>
        <title>HAL TEI export of hal-03805880</title>
      </titleStmt>
      <publicationStmt>
        <distributor>CCSD</distributor>
        <availability status="restricted">
          <licence target="https://creativecommons.org/publicdomain/zero/1.0/">CC0 1.0 - Universal</licence>
        </availability>
        <date when="2026-05-17T06:42:38+02:00"/>
      </publicationStmt>
      <sourceDesc>
        <p part="N">HAL API Platform</p>
      </sourceDesc>
    </fileDesc>
  </teiHeader>
  <text>
    <body>
      <listBibl>
        <biblFull>
          <titleStmt>
            <title xml:lang="en">Neural Networks with Physics-Informed Architectures and Constraints for Dynamical Systems Modeling</title>
            <author role="aut">
              <persName>
                <forename type="first">Franck</forename>
                <surname>Djeumou</surname>
              </persName>
              <email type="md5">a18d8d8ca45f4cc069001e173fc7981e</email>
              <email type="domain">utexas.edu</email>
              <idno type="idhal" notation="numeric">1172029</idno>
              <idno type="halauthorid" notation="string">2602609-1172029</idno>
              <affiliation ref="#struct-74980"/>
            </author>
            <author role="aut">
              <persName>
                <forename type="first">Cyrus</forename>
                <surname>Neary</surname>
              </persName>
              <email type="md5">4489793b20862d6d4dfadc8bacf95593</email>
              <email type="domain">utexas.edu</email>
              <idno type="idhal" notation="numeric">1172030</idno>
              <idno type="halauthorid" notation="string">2605758-1172030</idno>
              <affiliation ref="#struct-74980"/>
            </author>
            <author role="aut">
              <persName>
                <forename type="first">Eric</forename>
                <surname>Goubault</surname>
              </persName>
              <email type="md5">34691a3c485cac901a2837195944f0e2</email>
              <email type="domain">lix.polytechnique.fr</email>
              <idno type="idhal" notation="string">eric-goubault</idno>
              <idno type="idhal" notation="numeric">1172095</idno>
              <idno type="halauthorid" notation="string">208888-1172095</idno>
              <idno type="ORCID">https://orcid.org/0000-0002-3198-1863</idno>
              <affiliation ref="#struct-2071"/>
            </author>
            <author role="aut">
              <persName>
                <forename type="first">Sylvie</forename>
                <surname>Putot</surname>
              </persName>
              <email type="md5">623d4bf532ce4189123d269bca70b26b</email>
              <email type="domain">lix.polytechnique.fr</email>
              <idno type="idhal" notation="string">sylvie-putot</idno>
              <idno type="idhal" notation="numeric">184976</idno>
              <idno type="halauthorid" notation="string">119451-184976</idno>
              <idno type="ORCID">https://orcid.org/0000-0001-5624-3755</idno>
              <idno type="IDREF">https://www.idref.fr/181622769</idno>
              <affiliation ref="#struct-2071"/>
            </author>
            <author role="aut">
              <persName>
                <forename type="first">Ufuk</forename>
                <surname>Topcu</surname>
              </persName>
              <email type="md5">a4641ac8a07df2072a881c776b6e9739</email>
              <email type="domain">utexas.edu</email>
              <idno type="idhal" notation="numeric">1172031</idno>
              <idno type="halauthorid" notation="string">2602611-1172031</idno>
              <affiliation ref="#struct-74980"/>
            </author>
            <editor role="depositor">
              <persName>
                <forename>Sylvie</forename>
                <surname>Putot</surname>
              </persName>
              <email type="md5">623d4bf532ce4189123d269bca70b26b</email>
              <email type="domain">lix.polytechnique.fr</email>
            </editor>
          </titleStmt>
          <editionStmt>
            <edition n="v1" type="current">
              <date type="whenSubmitted">2022-10-12 15:09:24</date>
              <date type="whenModified">2024-08-01 17:08:02</date>
              <date type="whenReleased">2022-10-12 15:09:24</date>
              <date type="whenProduced">2022</date>
            </edition>
            <respStmt>
              <resp>contributor</resp>
              <name key="436938">
                <persName>
                  <forename>Sylvie</forename>
                  <surname>Putot</surname>
                </persName>
                <email type="md5">623d4bf532ce4189123d269bca70b26b</email>
                <email type="domain">lix.polytechnique.fr</email>
              </name>
            </respStmt>
          </editionStmt>
          <publicationStmt>
            <distributor>CCSD</distributor>
            <idno type="halId">hal-03805880</idno>
            <idno type="halUri">https://hal.science/hal-03805880</idno>
            <idno type="halBibtex">djeumou:hal-03805880</idno>
            <idno type="halRefHtml">&lt;i&gt;Learning for Dynamics and Control Conference,&lt;/i&gt;, 168, 2022, Proceedings of Machine Learning Research</idno>
            <idno type="halRef">Learning for Dynamics and Control Conference,, 168, 2022, Proceedings of Machine Learning Research</idno>
            <availability status="restricted"/>
          </publicationStmt>
          <seriesStmt>
            <idno type="stamp" n="X">École polytechnique</idno>
            <idno type="stamp" n="CNRS">CNRS - Centre national de la recherche scientifique</idno>
            <idno type="stamp" n="LIX" corresp="X">Laboratoire d'informatique de l'Ecole polytechnique</idno>
            <idno type="stamp" n="X-DEP-INFO">Département d'informatique de l’École polytechnique</idno>
            <idno type="stamp" n="IP_PARIS">Institut Polytechnique de Paris</idno>
          </seriesStmt>
          <notesStmt>
            <note type="audience" n="2">International</note>
            <note type="popular" n="0">No</note>
          </notesStmt>
          <sourceDesc>
            <biblStruct>
              <analytic>
                <title xml:lang="en">Neural Networks with Physics-Informed Architectures and Constraints for Dynamical Systems Modeling</title>
                <author role="aut">
                  <persName>
                    <forename type="first">Franck</forename>
                    <surname>Djeumou</surname>
                  </persName>
                  <email type="md5">a18d8d8ca45f4cc069001e173fc7981e</email>
                  <email type="domain">utexas.edu</email>
                  <idno type="idhal" notation="numeric">1172029</idno>
                  <idno type="halauthorid" notation="string">2602609-1172029</idno>
                  <affiliation ref="#struct-74980"/>
                </author>
                <author role="aut">
                  <persName>
                    <forename type="first">Cyrus</forename>
                    <surname>Neary</surname>
                  </persName>
                  <email type="md5">4489793b20862d6d4dfadc8bacf95593</email>
                  <email type="domain">utexas.edu</email>
                  <idno type="idhal" notation="numeric">1172030</idno>
                  <idno type="halauthorid" notation="string">2605758-1172030</idno>
                  <affiliation ref="#struct-74980"/>
                </author>
                <author role="aut">
                  <persName>
                    <forename type="first">Eric</forename>
                    <surname>Goubault</surname>
                  </persName>
                  <email type="md5">34691a3c485cac901a2837195944f0e2</email>
                  <email type="domain">lix.polytechnique.fr</email>
                  <idno type="idhal" notation="string">eric-goubault</idno>
                  <idno type="idhal" notation="numeric">1172095</idno>
                  <idno type="halauthorid" notation="string">208888-1172095</idno>
                  <idno type="ORCID">https://orcid.org/0000-0002-3198-1863</idno>
                  <affiliation ref="#struct-2071"/>
                </author>
                <author role="aut">
                  <persName>
                    <forename type="first">Sylvie</forename>
                    <surname>Putot</surname>
                  </persName>
                  <email type="md5">623d4bf532ce4189123d269bca70b26b</email>
                  <email type="domain">lix.polytechnique.fr</email>
                  <idno type="idhal" notation="string">sylvie-putot</idno>
                  <idno type="idhal" notation="numeric">184976</idno>
                  <idno type="halauthorid" notation="string">119451-184976</idno>
                  <idno type="ORCID">https://orcid.org/0000-0001-5624-3755</idno>
                  <idno type="IDREF">https://www.idref.fr/181622769</idno>
                  <affiliation ref="#struct-2071"/>
                </author>
                <author role="aut">
                  <persName>
                    <forename type="first">Ufuk</forename>
                    <surname>Topcu</surname>
                  </persName>
                  <email type="md5">a4641ac8a07df2072a881c776b6e9739</email>
                  <email type="domain">utexas.edu</email>
                  <idno type="idhal" notation="numeric">1172031</idno>
                  <idno type="halauthorid" notation="string">2602611-1172031</idno>
                  <affiliation ref="#struct-74980"/>
                </author>
              </analytic>
              <monogr>
                <title level="m">Learning for Dynamics and Control Conference,</title>
                <imprint>
                  <biblScope unit="serie">Proceedings of Machine Learning Research</biblScope>
                  <biblScope unit="volume">168</biblScope>
                  <date type="datePub">2022</date>
                </imprint>
              </monogr>
            </biblStruct>
          </sourceDesc>
          <profileDesc>
            <langUsage>
              <language ident="en">English</language>
            </langUsage>
            <textClass>
              <keywords scheme="author">
                <term xml:lang="en">Physics-constrained learning</term>
                <term xml:lang="en">neural ordinary differential equations</term>
                <term xml:lang="en">nonlinear system identification</term>
                <term xml:lang="en">dynamical systems</term>
                <term xml:lang="en">Physics-constrained learning</term>
              </keywords>
              <classCode scheme="halDomain" n="info.info-ai">Computer Science [cs]/Artificial Intelligence [cs.AI]</classCode>
              <classCode scheme="halTypology" n="COUV">Book sections</classCode>
              <classCode scheme="halOldTypology" n="COUV">Book sections</classCode>
              <classCode scheme="halTreeTypology" n="COUV">Book sections</classCode>
            </textClass>
            <abstract xml:lang="en">
              <p>Effective inclusion of physics-based knowledge into deep neural network models of dynamical systems can greatly improve data efficiency and generalization. Such a priori knowledge might arise from physical principles (e.g., conservation laws) or from the system's design (e.g., the Jacobian matrix of a robot), even if large portions of the system dynamics remain unknown. We develop a framework to learn dynamics models from trajectory data while incorporating a priori system knowledge as inductive bias. More specifically, the proposed framework uses physics-based side information to inform the structure of the neural network itself, and to place constraints on the values of the outputs and the internal states of the model. It represents the system's vector field as a composition of known and unknown functions, the latter of which are parametrized by neural networks. The physics-informed constraints are enforced via the augmented Lagrangian method during the model's training. We experimentally demonstrate the benefits of the proposed approach on a variety of dynamical systems-including a benchmark suite of robotics environments featuring large state spaces, non-linear dynamics, external forces, contact forces, and control inputs. By exploiting a priori system knowledge during training, the proposed approach learns to predict the system dynamics two orders of magnitude more accurately than a baseline approach that does not include prior knowledge, given the same training dataset.</p>
            </abstract>
          </profileDesc>
        </biblFull>
      </listBibl>
    </body>
    <back>
      <listOrg type="structures">
        <org type="institution" xml:id="struct-74980" status="VALID">
          <idno type="ROR">https://ror.org/00hj54h04</idno>
          <orgName>University of Texas at Austin [Austin]</orgName>
          <desc>
            <address>
              <addrLine>1 University Station Austin, Texas 78712</addrLine>
              <country key="US"/>
            </address>
            <ref type="url">http://www.utexas.edu/</ref>
          </desc>
        </org>
        <org type="laboratory" xml:id="struct-2071" status="VALID">
          <idno type="IdRef">196509955</idno>
          <idno type="ISNI">0000000403673665</idno>
          <idno type="RNSR">200519331V</idno>
          <idno type="ROR">https://ror.org/04afed728</idno>
          <idno type="Wikidata">Q16009025</idno>
          <orgName>Laboratoire d'informatique de l'École polytechnique [Palaiseau]</orgName>
          <orgName type="acronym">LIX</orgName>
          <date type="start">2005-01-01</date>
          <desc>
            <address>
              <addrLine>1 Rue Honoré d’Estienne d’Orves, Bâtiment Alain Turing, 91120 Palaiseau</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">https://www.lix.polytechnique.fr/</ref>
          </desc>
          <listRelation>
            <relation active="#struct-300340" type="direct"/>
            <relation active="#struct-563936" type="indirect"/>
            <relation name="UMR7161" active="#struct-441569" type="direct"/>
          </listRelation>
        </org>
        <org type="institution" xml:id="struct-300340" status="VALID">
          <idno type="IdRef">027309320</idno>
          <idno type="ISNI">0000000121581279</idno>
          <idno type="ROR">https://ror.org/05hy3tk52</idno>
          <idno type="Wikidata">Q273626</idno>
          <orgName>École polytechnique</orgName>
          <orgName type="acronym">X</orgName>
          <date type="start">1794-03-11</date>
          <desc>
            <address>
              <addrLine>Route de Saclay, 91128 Palaiseau Cedex</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">https://www.polytechnique.edu/</ref>
          </desc>
          <listRelation>
            <relation active="#struct-563936" type="direct"/>
          </listRelation>
        </org>
        <org type="regroupinstitution" xml:id="struct-563936" status="VALID">
          <idno type="IdRef">238327159</idno>
          <idno type="ISNI">0000000502717600</idno>
          <idno type="ROR">https://ror.org/042tfbd02</idno>
          <idno type="Wikidata">Q48759778</idno>
          <orgName>Institut Polytechnique de Paris</orgName>
          <orgName type="acronym">IP Paris</orgName>
          <date type="start">2019-06-02</date>
          <desc>
            <address>
              <addrLine>Route de Saclay, 91120 Palaiseau Cedex, France</addrLine>
              <country key="FR"/>
            </address>
            <ref type="url">https://www.ip-paris.fr</ref>
          </desc>
        </org>
        <org type="regroupinstitution" xml:id="struct-441569" status="VALID">
          <idno type="IdRef">02636817X</idno>
          <idno type="ISNI">0000000122597504</idno>
          <idno type="ROR">https://ror.org/02feahw73</idno>
          <orgName>Centre National de la Recherche Scientifique</orgName>
          <orgName type="acronym">CNRS</orgName>
          <date type="start">1939-10-19</date>
          <desc>
            <address>
              <country key="FR"/>
            </address>
            <ref type="url">https://www.cnrs.fr/</ref>
          </desc>
        </org>
      </listOrg>
    </back>
  </text>
</TEI>