A fast time-stepping strategy for dynamical systems equipped with a surrogate model
From MaRDI portal
Publication:5864692
DOI10.1137/20M1386281zbMATH Open1492.65196arXiv2011.03688OpenAlexW3112623379MaRDI QIDQ5864692FDOQ5864692
Authors: Steven Roberts, Andrey Popov, Arash Sarshar, Adrian Sandu 
Publication date: 8 June 2022
Published in: SIAM Journal on Scientific Computing (Search for Journal in Brave)
Abstract: Simulation of complex dynamical systems arising in many applications is computationally challenging due to their size and complexity. Model order reduction, machine learning, and other types of surrogate modeling techniques offer cheaper and simpler ways to describe the dynamics of these systems but are inexact and introduce additional approximation errors. In order to overcome the computational difficulties of the full complex models, on one hand, and the limitations of surrogate models, on the other, this work proposes a new accelerated time-stepping strategy that combines information from both. This approach is based on the multirate infinitesimal general-structure additive Runge-Kutta (MRI-GARK) framework. The inexpensive surrogate model is integrated with a small timestep to guide the solution trajectory, and the full model is treated with a large timestep to occasionally correct for the surrogate model error and ensure convergence. We provide a theoretical error analysis, and several numerical experiments, to show that this approach can be significantly more efficient than using only the full or only the surrogate model for the integration.
Full work available at URL: https://arxiv.org/abs/2011.03688
Recommendations
- Projection-based model reduction of dynamical systems using space-time subspace and machine learning
- A data-driven surrogate modeling approach for time-dependent incompressible Navier-Stokes equations with dynamic mode decomposition and manifold interpolation
- Non-intrusive model reduction of large-scale, nonlinear dynamical systems using deep learning
- Time-series machine-learning error models for approximate solutions to parameterized dynamical systems
- Decomposition-assisted computational technique based on surrogate modeling for real-time simulations
machine learningreduced-order modelingsurrogate modelmultirate time integrationgeneral-structure additive Runge-Kutta methods
Multistep, Runge-Kutta and extrapolation methods for ordinary differential equations (65L06) Error bounds for numerical methods for ordinary differential equations (65L70) Numerical methods for initial value problems involving ordinary differential equations (65L05)
Cites Work
- MFEM: a modular finite element methods library
- Dynamic mode decomposition of numerical and experimental data
- Multirate timestepping methods for hyperbolic conservation laws
- Nonlinear model reduction via discrete empirical interpolation
- Solving Ordinary Differential Equations I
- Extrapolation-based implicit-explicit general linear methods
- High order implicit-explicit general linear methods with optimized stability regions
- A Class Of Implicit-Explicit Two-Step Runge--Kutta Methods
- Turbulence and the dynamics of coherent structures. I. Coherent structures
- Partitioned and implicit-explicit general linear methods for ordinary differential equations
- Extrapolated multirate methods for differential equations with multiple time scales
- The Reduced Basis Method for Incompressible Viscous Flow Calculations
- Numerical Solution of Systems of Ordinary Differential Equations Separated into Subsystems
- Multirate explicit Adams methods for time integration of conservation laws
- Multirate linear multistep methods
- The heterogeneous multiscale method
- Implicit-explicit Runge-Kutta methods for computing atmospheric reactive flows
- A finite element discretization of the streamfunction formulation of the stationary quasi-geostrophic equations of the Ocean
- Computational design for long-term numerical integration of the equations of fluid motion: two-dimensional incompressible flow. Part I
- Multirate Runge-Kutta schemes for advection equations
- A new class of exponential propagation iterative methods of Runge-Kutta type (EPIRK)
- On HMM-like integrators and projective integration methods for systems with multiple time scales
- Symplectic Methods Based on Decompositions
- Multirate infinitesimal step methods for atmospheric flow simulation
- Runge-Kutta Methods with Minimum Error Bounds
- Split Runge-Kutta method for simultaneous equations
- Multirate generalized additive Runge Kutta methods
- Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
- Data-driven science and engineering. Machine learning, dynamical systems, and control
- Implicit multirate GARK methods
- A stabilized proper orthogonal decomposition reduced-order model for large scale quasigeostrophic ocean circulation
- EPIRK-\(W\) and EPIRK-\(K\) time discretization methods
- Multiscale high-order/low-order (HOLO) algorithms and applications
- Multirate implicit Euler schemes for a class of differential-algebraic equations of index-1
- Design of high-order decoupled multirate GARK schemes
- Coupled Multirate Infinitesimal GARK Schemes for Stiff Systems with Multiple Time Scales
- A class of multirate infinitesimal GARK methods
- A conforming finite element discretization of the streamfunction form of the unsteady quasi-geostrophic equations
- Data-driven correction reduced order models for the quasi-geostrophic equations: a numerical investigation
- A multifidelity ensemble Kalman filter with reduced order control variates
- Linearly implicit GARK schemes
- Convergence Acceleration for Time-Dependent Parametric Multifidelity Models
- Extended multirate infinitesimal step methods: derivation of order conditions
Cited In (4)
- Deterministic and stochastic surrogate models for a slowly driven fast oscillator
- Decomposition-assisted computational technique based on surrogate modeling for real-time simulations
- Adaptive learning of effective dynamics for online modeling of complex systems
- Surrogate Models for Oscillatory Systems Using Sparse Polynomial Chaos Expansions and Stochastic Time Warping
Uses Software
This page was built for publication: A fast time-stepping strategy for dynamical systems equipped with a surrogate model
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5864692)
