Learning in Modal Space: Solving Time-Dependent Stochastic PDEs Using Physics-Informed Neural Networks
From MaRDI portal
Publication:5221026
Abstract: One of the open problems in scientific computing is the long-time integration of nonlinear stochastic partial differential equations (SPDEs). We address this problem by taking advantage of recent advances in scientific machine learning and the dynamically orthogonal (DO) and bi-orthogonal (BO) methods for representing stochastic processes. Specifically, we propose two new Physics-Informed Neural Networks (PINNs) for solving time-dependent SPDEs, namely the NN-DO/BO methods, which incorporate the DO/BO constraints into the loss function with an implicit form instead of generating explicit expressions for the temporal derivatives of the DO/BO modes. Hence, the proposed methods overcome some of the drawbacks of the original DO/BO methods: we do not need the assumption that the covariance matrix of the random coefficients is invertible as in the original DO method, and we can remove the assumption of no eigenvalue crossing as in the original BO method. Moreover, the NN-DO/BO methods can be used to solve time-dependent stochastic inverse problems with the same formulation and computational complexity as for forward problems. We demonstrate the capability of the proposed methods via several numerical examples: (1) A linear stochastic advection equation with deterministic initial condition where the original DO/BO method would fail; (2) Long-time integration of the stochastic Burgers' equation with many eigenvalue crossings during the whole time evolution where the original BO method fails. (3) Nonlinear reaction diffusion equation: we consider both the forward and the inverse problem, including noisy initial data, to investigate the flexibility of the NN-DO/BO methods in handling inverse and mixed type problems. Taken together, these simulation results demonstrate that the NN-DO/BO methods can be employed to effectively quantify uncertainty propagation in a wide range of physical problems.
Recommendations
- Data-driven deep learning of partial differential equations in modal space
- Stochastic physics-informed neural ordinary differential equations
- Deep physics corrector: a physics enhanced deep learning architecture for solving stochastic differential equations
- A deep learning approach to the probabilistic numerical solution of path-dependent partial differential equations
- Learning time-dependent PDEs with a linear and nonlinear separate convolutional neural network
- Neural-PDE: a RNN based neural network for solving time dependent PDEs
- Modeling the dynamics of PDE systems with physics-constrained deep auto-regressive networks
- Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
- Physics-informed generative adversarial networks for stochastic differential equations
- Deep hidden physics models: deep learning of nonlinear partial differential equations
Cites work
- A ``parareal in time discretization of PDE's
- A convergence study for SPDEs using combined polynomial chaos and dynamically-orthogonal schemes
- A dynamically bi-orthogonal method for time-dependent stochastic partial differential equations. I: Derivation and algorithms
- A dynamically bi-orthogonal method for time-dependent stochastic partial differential equations. II: Adaptivity and generalizations
- A robust bi-orthogonal/dynamically-orthogonal method using the covariance pseudo-inverse with application to stochastic flow problems
- Approximation by superpositions of a sigmoidal function
- Deep learning of vortex-induced vibrations
- Dynamical criteria for the evolution of the stochastic dimensionality in flows with uncertainty
- Dynamically orthogonal field equations for continuous stochastic dynamical systems
- Error analysis of the dynamically orthogonal approximation of time dependent random PDEs
- Hidden physics models: machine learning of nonlinear partial differential equations
- High-Order Collocation Methods for Differential Equations with Random Inputs
- Inferring solutions of differential equations using noisy multi-fidelity data
- Numerical methods for stochastic computations. A spectral method approach.
- On the equivalence of dynamically orthogonal and bi-orthogonal methods: theory and numerical simulations
- Quantifying total uncertainty in physics-informed neural networks for solving forward and inverse stochastic problems
- Stochastic collocation on unstructured multivariate meshes
- The Wiener--Askey Polynomial Chaos for Stochastic Differential Equations
Cited in
(only showing first 100 items - show all)- PPINN: parareal physics-informed neural network for time-dependent PDEs
- A physics-informed operator regression framework for extracting data-driven continuum models
- A multigrid multilevel Monte Carlo method for Stokes-Darcy model with random hydraulic conductivity and Beavers-Joseph condition
- PhyCRNet: physics-informed convolutional-recurrent network for solving spatiotemporal PDEs
- Data-driven peakon and periodic peakon solutions and parameter discovery of some nonlinear dispersive equations via deep learning
- Data-driven discoveries of Bäcklund transformations and soliton evolution equations via deep neural network learning schemes
- Optimal control of PDEs using physics-informed neural networks
- A minimalistic approach to physics-informed machine learning using neighbour lists as physics-optimized convolutions for inverse problems involving particle systems
- Learning stochastic dynamics with statistics-informed neural network
- Data-driven deep learning of partial differential equations in modal space
- On some neural network architectures that can represent viscosity solutions of certain high dimensional Hamilton-Jacobi partial differential equations
- Structure-preserving neural networks
- A stochastic collocation method based on sparse grids for a stochastic Stokes-Darcy model
- Meta-mgnet: meta multigrid networks for solving parameterized partial differential equations
- Thermodynamically consistent physics-informed neural networks for hyperbolic systems
- When and why PINNs fail to train: a neural tangent kernel perspective
- A neural network multigrid solver for the Navier-Stokes equations
- A comprehensive and fair comparison of two neural operators (with practical extensions) based on FAIR data
- Gradient-enhanced physics-informed neural networks for forward and inverse PDE problems
- Learning generative neural networks with physics knowledge
- Mean-field and kinetic descriptions of neural differential equations
- Scientific machine learning through physics-informed neural networks: where we are and what's next
- Physics-informed PointNet: a deep learning solver for steady-state incompressible flows and thermal fields on multiple sets of irregular geometries
- HomPINNs: Homotopy physics-informed neural networks for learning multiple solutions of nonlinear elliptic differential equations
- Quantifying total uncertainty in physics-informed neural networks for solving forward and inverse stochastic problems
- Simulator-free solution of high-dimensional stochastic elliptic partial differential equations using deep neural networks
- Data-driven vector soliton solutions of coupled nonlinear Schrödinger equation using a deep learning algorithm
- A-PINN: auxiliary physics informed neural networks for forward and inverse problems of nonlinear integro-differential equations
- A-WPINN algorithm for the data-driven vector-soliton solutions and parameter discovery of general coupled nonlinear equations
- Data-driven spatiotemporal modeling for structural dynamics on irregular domains by stochastic dependency neural estimation
- Uncertainty quantification in scientific machine learning: methods, metrics, and comparisons
- Surrogate modeling for Bayesian inverse problems based on physics-informed neural networks
- Neural network architectures using min-plus algebra for solving certain high-dimensional optimal control problems and Hamilton-Jacobi PDEs
- An overview on deep learning-based approximation methods for partial differential equations
- Learning and meta-learning of stochastic advection–diffusion–reaction systems from sparse measurements
- Algorithms for solving high dimensional PDEs: from nonlinear Monte Carlo to machine learning
- An Augmented Lagrangian Deep Learning Method for Variational Problems with Essential Boundary Conditions
- DeepXDE: a deep learning library for solving differential equations
- Extended Physics-Informed Neural Networks (XPINNs): A Generalized Space-Time Domain Decomposition Based Deep Learning Framework for Nonlinear Partial Differential Equations
- On the convergence of physics informed neural networks for linear second-order elliptic and parabolic type PDEs
- Deep Nitsche Method: Deep Ritz Method with Essential Boundary Conditions
- Physics-informed neural networks with hard constraints for inverse design
- Physics-informed generative adversarial networks for stochastic differential equations
- Randomized neural network with Petrov-Galerkin methods for solving linear and nonlinear partial differential equations
- Application of the dynamical system method and the deep learning method to solve the new (3+1)-dimensional fractional modified Benjamin-Bona-Mahony equation
- VC-PINN: variable coefficient physics-informed neural network for forward and inverse problems of PDEs with variable coefficient
- Complex dynamics on the one-dimensional quantum droplets via time piecewise PINNs
- Continuous limits of residual neural networks in case of large input data
- FDM-PINN: Physics-informed neural network based on fictitious domain method
- Pre-training strategy for solving evolution equations based on physics-informed neural networks
- Pseudo-Hamiltonian neural networks for learning partial differential equations
- Bi-Orthogonal fPINN: A Physics-Informed Neural Network Method for Solving Time-Dependent Stochastic Fractional PDEs
- Bayesian Deep Learning Framework for Uncertainty Quantification in Stochastic Partial Differential Equations
- Asymptotic-Preserving Neural Networks for multiscale hyperbolic models of epidemic spread
- Physics-informed variational inference for uncertainty quantification of stochastic differential equations
- Posteriori error neural network: a recovery type posteriori error estimator based on neural network for diffusion problems
- AONN: An Adjoint-Oriented Neural Network Method for All-At-Once Solutions of Parametric Optimal Control Problems
- Roenet: predicting discontinuity of hyperbolic systems from continuous data
- Splitting physics-informed neural networks for inferring the dynamics of integer- and fractional-order neuron models
- Adaptive deep neural networks for solving corner singular problems
- Solving the non-local Fokker-Planck equations by deep learning
- CEENs: causality-enforced evolutional networks for solving time-dependent partial differential equations
- Neuro-PINN: a hybrid framework for efficient nonlinear projection equation solutions
- Bright-dark rogue wave transition in coupled ab system via the physics-informed neural networks method
- A deep learning method for solving multi-dimensional coupled forward-backward doubly SDEs
- Solving large-scale variational inequalities with dynamically adjusting initial condition in physics-informed neural networks
- E-PINN: extended physics informed neural network for the forward and inverse problems of high-order nonlinear integro-differential equations
- Machine learning for nonlinear integro-differential equations with degenerate kernel scheme
- Non-intrusive reduced-order model for time-dependent stochastic partial differential equations utilizing dynamic mode decomposition and polynomial chaos expansion
- Asymptotic-preserving neural networks for hyperbolic systems with diffusive scaling
- Deep surrogate model for learning Green's function associated with linear reaction-diffusion operator
- Prediction of spatiotemporal dynamics using deep learning: coupled neural networks of long short-terms memory, auto-encoder and physics-informed neural networks
- Time integration schemes based on neural networks for solving partial differential equations on coarse grids
- Application of PINN to obtain solution of boundary layer problems arising in fluid dynamics
- Investigation of forward and inverse problems in a nonlocal reverse-time nonlinear Schrödinger equation using physics-informed neural networks
- Multi-benchmark adaptive sampling physics-informed neural network for complex and coupled equations
- Reconstruction and prediction of Volterra integral equations driven by Gaussian noise
- Nonlinear interaction multi-domain physics-informed neural networks for the hybrid rogue wave and breather solutions of the nonlinear Schrödinger equation
- RTC-fPINNs: respecting temporal causality fPINNs method for time fractional fourth order partial differential equations
- Solving Lane-Emden-type eigenvalue problems with physics-informed neural networks
- On the regularity theory for quasilinear elliptic systems with the application of Leray-Schauder method
- Data-driven SFB solutions and parameters discovery for nonlinear Schrödinger equation via time domain decomposition physics-informed neural networks
- Multilevel Picard approximations overcome the curse of dimensionality when approximating semilinear heat equations with gradient-dependent nonlinearities in L^p-sense
- Reduced-order modeling for uncertainty quantification of parameterized fluid-structure interaction problems
- A natural deep Ritz method for essential boundary value problems
- A novel deep convolutional surrogate model with incomplete solve loss for parameterized steady-state diffusion problems
- Regularity of the solutions to quasi-linear parabolic systems with the singular coefficients
- Multilevel Picard approximations overcome the curse of dimensionality in the numerical approximation of general semilinear PDEs with gradient-dependent nonlinearities
- Physics-informed neural networks (PINNs) for fluid mechanics: a review
- Full error analysis of the random deep splitting method for nonlinear parabolic PDEs and PIDEs
- Extended physics-informed extreme learning machine for linear elastic fracture mechanics
- Approximation of solution to differential systems driven by Rosenblatt process containing noninstantaneous impulses
- Solving multi-group neutron diffusion eigenvalue problem with decoupling residual loss function
- Physics-informed multiresolution wavelet neural network method for solving partial differential equations
- Convolution-weighting method for the physics-informed neural network: a primal-dual optimization perspective
- Kernel-based PINNs for time fractional differential equations and Volterra integral-differential equations
- Multilevel Picard algorithm for general semilinear parabolic PDEs with gradient-dependent nonlinearities
- A novel adaptive gradient-enhanced physics-informed neural networks for solving partial differential equations
- Physics-informed neural networks for PDE-constrained optimization and control
- Self-attention enhanced physics-informed neural networks for multi-field physics
This page was built for publication: Learning in Modal Space: Solving Time-Dependent Stochastic PDEs Using Physics-Informed Neural Networks
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5221026)
