Constructing Neural Network Based Models for Simulating Dynamical Systems

Christian Møldrup Legaard; Thomas Schranz; Gerald Schweiger; Jan Drgona; Basak Falay; Cláudio Gomes; Alexandros Iosifidis; Mahdi Abkar; Peter Gorm Larsen

doi:10.1145/3567591

Constructing Neural Network Based Models for Simulating Dynamical Systems

Christian Møldrup Legaard, Thomas Schranz, Gerald Schweiger, Jan Drgona, Basak Falay, Cláudio Gomes, Alexandros Iosifidis, Mahdi Abkar, Peter Gorm Larsen

Publikation: Bidrag til tidsskrift/Konferencebidrag i tidsskrift /Bidrag til avis › Tidsskriftartikel › Forskning › peer review

41 Citationer (Scopus)

Abstract

Dynamical systems see widespread use in natural sciences like physics, biology, and chemistry, as well as engineering disciplines such as circuit analysis, computational fluid dynamics, and control. For simple systems, the differential equations governing the dynamics can be derived by applying fundamental physical laws. However, for more complex systems, this approach becomes exceedingly difficult. Data-driven modeling is an alternative paradigm that seeks to learn an approximation of the dynamics of a system using observations of the true system. In recent years, there has been an increased interest in applying data-driven modeling techniques to solve a wide range of problems in physics and engineering. This article provides a survey of the different ways to construct models of dynamical systems using neural networks. In addition to the basic overview, we review the related literature and outline the most significant challenges from numerical simulations that this modeling paradigm must overcome. Based on the reviewed literature and identified challenges, we provide a discussion on promising research areas.

Originalsprog	Engelsk
Artikelnummer	236
Tidsskrift	ACM Computing Surveys
Vol/bind	55
Nummer	11
Sider (fra-til)	1-34
Antal sider	34
ISSN	0360-0300
DOI	https://doi.org/10.1145/3567591
Status	Udgivet - feb. 2023

Adgang til dokumentet

10.1145/3567591

https://arxiv.org/pdf/2111.01495Licens: Ikke-specificeret

Andre filer og links

Link to publication in Scopus

Citationsformater

@article{e73c1ef890474df39ba0d66136915029,

title = "Constructing Neural Network Based Models for Simulating Dynamical Systems",

abstract = "Dynamical systems see widespread use in natural sciences like physics, biology, and chemistry, as well as engineering disciplines such as circuit analysis, computational fluid dynamics, and control. For simple systems, the differential equations governing the dynamics can be derived by applying fundamental physical laws. However, for more complex systems, this approach becomes exceedingly difficult. Data-driven modeling is an alternative paradigm that seeks to learn an approximation of the dynamics of a system using observations of the true system. In recent years, there has been an increased interest in applying data-driven modeling techniques to solve a wide range of problems in physics and engineering. This article provides a survey of the different ways to construct models of dynamical systems using neural networks. In addition to the basic overview, we review the related literature and outline the most significant challenges from numerical simulations that this modeling paradigm must overcome. Based on the reviewed literature and identified challenges, we provide a discussion on promising research areas.",

keywords = "Additional Key Words and PhrasesNeural ODEs, physics-based regularization, physics-informed neural networks",

author = "Legaard, {Christian M{\o}ldrup} and Thomas Schranz and Gerald Schweiger and Jan Drgona and Basak Falay and Cl{\'a}udio Gomes and Alexandros Iosifidis and Mahdi Abkar and Larsen, {Peter Gorm}",

year = "2023",

month = feb,

doi = "10.1145/3567591",

language = "English",

volume = "55",

pages = "1--34",

journal = "ACM Computing Surveys",

issn = "0360-0300",

publisher = "ACM",

number = "11",

}

TY - JOUR

T1 - Constructing Neural Network Based Models for Simulating Dynamical Systems

AU - Legaard, Christian Møldrup

AU - Schranz, Thomas

AU - Schweiger, Gerald

AU - Drgona, Jan

AU - Falay, Basak

AU - Gomes, Cláudio

AU - Iosifidis, Alexandros

AU - Abkar, Mahdi

AU - Larsen, Peter Gorm

PY - 2023/2

Y1 - 2023/2

N2 - Dynamical systems see widespread use in natural sciences like physics, biology, and chemistry, as well as engineering disciplines such as circuit analysis, computational fluid dynamics, and control. For simple systems, the differential equations governing the dynamics can be derived by applying fundamental physical laws. However, for more complex systems, this approach becomes exceedingly difficult. Data-driven modeling is an alternative paradigm that seeks to learn an approximation of the dynamics of a system using observations of the true system. In recent years, there has been an increased interest in applying data-driven modeling techniques to solve a wide range of problems in physics and engineering. This article provides a survey of the different ways to construct models of dynamical systems using neural networks. In addition to the basic overview, we review the related literature and outline the most significant challenges from numerical simulations that this modeling paradigm must overcome. Based on the reviewed literature and identified challenges, we provide a discussion on promising research areas.

AB - Dynamical systems see widespread use in natural sciences like physics, biology, and chemistry, as well as engineering disciplines such as circuit analysis, computational fluid dynamics, and control. For simple systems, the differential equations governing the dynamics can be derived by applying fundamental physical laws. However, for more complex systems, this approach becomes exceedingly difficult. Data-driven modeling is an alternative paradigm that seeks to learn an approximation of the dynamics of a system using observations of the true system. In recent years, there has been an increased interest in applying data-driven modeling techniques to solve a wide range of problems in physics and engineering. This article provides a survey of the different ways to construct models of dynamical systems using neural networks. In addition to the basic overview, we review the related literature and outline the most significant challenges from numerical simulations that this modeling paradigm must overcome. Based on the reviewed literature and identified challenges, we provide a discussion on promising research areas.

KW - Additional Key Words and PhrasesNeural ODEs

KW - physics-based regularization

KW - physics-informed neural networks

UR - http://www.scopus.com/inward/record.url?scp=85151845803&partnerID=8YFLogxK

U2 - 10.1145/3567591

DO - 10.1145/3567591

M3 - Journal article

SN - 0360-0300

VL - 55

SP - 1

EP - 34

JO - ACM Computing Surveys

JF - ACM Computing Surveys

IS - 11

M1 - 236

ER -

Constructing Neural Network Based Models for Simulating Dynamical Systems

Abstract

Adgang til dokumentet

Andre filer og links

Fingeraftryk

Citationsformater