r-adaptive deep learning method for solving partial differential equations
From MaRDI portal
\(r\)-adaptive deep learning method for solving partial differential equations
Mesh generation, refinement, and adaptive methods for the numerical solution of initial value and initial-boundary value problems involving PDEs (65M50) Error bounds for boundary value problems involving PDEs (65N15) Finite element, Rayleigh-Ritz and Galerkin methods for boundary value problems involving PDEs (65N30) Learning and adaptive systems in artificial intelligence (68T05) Artificial neural networks and deep learning (68T07)
Abstract: We introduce an adaptive algorithm to solve Partial Differential Equations using a Deep Neural Network. The proposed method restricts to tensor product meshes and optimizes the boundary node locations in one dimension, from which we build two- or three-dimensional meshes. The method allows the definition of fixed interfaces to design conforming meshes, and enables changes in the topology, i.e., some nodes can jump across fixed interfaces. The method simultaneously optimizes the node locations and the PDE solution values over the resulting mesh. To numerically illustrate the performance of our proposed adaptive method, we apply it in combination with a collocation method, a Least Squares Method, and a Deep Ritz Method. We focus on the latter to solve one- and two-dimensional problems whose solutions are smooth, singular, and/or exhibit strong gradients.
Recommendations
- Adaptive deep neural networks methods for high-dimensional partial differential equations
- A deep double Ritz method (\(\mathrm{D^2RM}\)) for solving partial differential equations using neural networks
- DAS-PINNs: a deep adaptive sampling method for solving high-dimensional partial differential equations
- The deep Ritz method: a deep learning-based numerical algorithm for solving variational problems
Cites work
- A finite element based deep learning solver for parametric PDEs
- A machine-learning minimal-residual (ML-MRes) framework for goal-oriented finite element discretizations
- Adaptive moving mesh methods
- Adaptivity with moving grids
- Algorithms for solving high dimensional PDEs: from nonlinear Monte Carlo to machine learning
- Analysis of Optimal Finite-Element Meshes in R 1
- Constrained Delaunay triangulations
- Deep learning
- Deep learning in computational mechanics. An introductory course
- Deep learning: methods and applications
- Deep least-squares methods: an unsupervised learning-based numerical method for solving elliptic PDEs
- DGM: a deep learning algorithm for solving partial differential equations
- Evaluating Derivatives
- Finite cell method. \(h\)- and \(p\)-extension for embedded domain problems in solid mechanics
- Fundamentals of Engineering Numerical Analysis
- scientific article; zbMATH DE number 42145 (Why is no real title available?)
- Learning representations by back-propagating errors
- Mesh Generation
- Moving Mesh Partial Differential Equations (MMPDES) Based on the Equidistribution Principle
- Neural control of discrete weak formulations: Galerkin, least squares \& minimal-residual methods with quasi-optimal weights
- On quadrature rules for solving partial differential equations using neural networks
- Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
- Piecewise polynomial approximation on optimal meshes
- Recurrent neural networks as optimal mesh refinement strategies
- Simple adaptive grids for 1-D initial value problems
- Solving high-dimensional partial differential equations using deep learning
- The deep Ritz method: a deep learning-based numerical algorithm for solving variational problems
- The finite cell method for linear thermoelasticity
- The finite element methods and its reliability
- The h, p and h-p versions of the finite element method in 1 dimension. II. The error analysis of the h- and h-p versions
- The scaling and skewness of optimally transported meshes on the sphere
Cited in
(9)- Adaptive deep neural networks methods for high-dimensional partial differential equations
- DAS-PINNs: a deep adaptive sampling method for solving high-dimensional partial differential equations
- A deep double Ritz method (\(\mathrm{D^2RM}\)) for solving partial differential equations using neural networks
- N-adaptive Ritz method: a neural network enriched partition of unity for boundary value problems
- A deep learning method for solving multi-dimensional coupled forward-backward doubly SDEs
- An r-adaptive finite element method using neural networks for parametric self-adjoint elliptic problems
- High-precision physics-informed extreme learning machines for evolving interface problems
- Finite element neural network interpolation. I: Interpretable and adaptive discretization for solving PDEs
- Stochastic quadrature rules for solving PDEs using neural networks
This page was built for publication: \(r\)-adaptive deep learning method for solving partial differential equations
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6144172)
