Simulating Partial Differential Equations with Neural Networks

被引:0
作者
Chertock, Anna [1 ]
Leonard, Christopher [1 ]
机构
[1] North Carolina State Univ, Raleigh, NC 27695 USA
来源
HYPERBOLIC PROBLEMS: THEORY, NUMERICS, APPLICATIONS, VOL II, HYP2022 | 2024年 / 35卷
关键词
Neural networks; Partial differential equations; Finite-volume methods; SCHEMES;
D O I
10.1007/978-3-031-55264-9_4
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
In this paper, we present a novel approach for simulating solutions of partial differential equations using neural networks. We consider a time-stepping method similar to the finite-volume method, where the flux terms are computed using neural networks. To train the neural network, we collect 'sensor' data on small subsets of the computational domain. Thus, our neural network learns the local behavior of the solution rather than the global one. This leads to a much more versatile method that can simulate the solution to equations whose initial conditions are not in the same form as the initial conditions we train with. Also, using sensor data from a small portion of the domain is much more realistic than methods where a neural network is trained using data over a large domain.
引用
收藏
页码:39 / 49
页数:11
相关论文
共 50 条
[41]   A Deep Double Ritz Method (D2RM) for solving Partial Differential Equations using Neural Networks [J].
Uriarte, Carlos ;
Pardo, David ;
Muga, Ignacio ;
Munoz-Matute, Judit .
COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING, 2023, 405
[42]   Trans-Net: A transferable pretrained neural networks based on temporal domain decomposition for solving partial differential equations [J].
Zhang, Dinglei ;
Li, Ying ;
Ying, Shihui .
COMPUTER PHYSICS COMMUNICATIONS, 2024, 299
[43]   Isogeometric neural networks: A new deep learning approach for solving parameterized partial differential equations [J].
Gasick, Joshua ;
Qian, Xiaoping .
COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING, 2023, 405
[44]   Neural network method for solving partial differential equations [J].
Aarts, LP ;
van der Veer, P .
NEURAL PROCESSING LETTERS, 2001, 14 (03) :261-271
[45]   Neural Network Method for Solving Partial Differential Equations [J].
Lucie P. Aarts ;
Peter van der Veer .
Neural Processing Letters, 2001, 14 :261-271
[46]   Solving differential equations with constructed neural networks [J].
Tsoulos, Ioannis G. ;
Gavrilis, Dimitris ;
Glavas, Euripidis .
NEUROCOMPUTING, 2009, 72 (10-12) :2385-2391
[47]   A trial solution for imposing boundary conditions of partial differential equations in physics-informed neural networks [J].
Manavi, Seyedalborz ;
Fattahi, Ehsan ;
Becker, Thomas .
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE, 2024, 127
[48]   Physics-informed neural networks combined with polynomial interpolation to solve nonlinear partial differential equations [J].
Tang, Siping ;
Feng, Xinlong ;
Wu, Wei ;
Xu, Hui .
COMPUTERS & MATHEMATICS WITH APPLICATIONS, 2023, 132 :48-62
[49]   DWNN: DeepWavelet Neural Network for Solving Partial Differential Equations [J].
Li, Ying ;
Xu, Longxiang ;
Ying, Shihui .
MATHEMATICS, 2022, 10 (12)
[50]   A comparative investigation of neural networks in solving differential equations [J].
Shi, Enze ;
Xu, Chuanju .
JOURNAL OF ALGORITHMS & COMPUTATIONAL TECHNOLOGY, 2021, 15