Abstract:Projection-based model reduction has become a popular approach to reduce the cost associated with integrating large-scale dynamical systems so they can be used in many-query settings such as optimization and uncertainty quantification. For nonlinear systems, significant cost reduction is only possible with an additional layer of approximation to reduce the computational bottleneck of evaluating the projected nonlinear terms. Prevailing methods to approximate the nonlinear terms are code intrusive, potentially requiring years of development time to integrate into an existing codebase, and have been known to lack parametric robustness. This work develops a non-intrusive method to efficiently and accurately approximate the expensive nonlinear terms that arise in reduced nonlinear dynamical system using deep neural networks. The neural network is trained using only the simulation data used to construct the reduced basis and evaluations of the nonlinear terms at these snapshots. Once trained, the neural network-based reduced-order model only requires forward and backward propagation through the network to evaluate the nonlinear term and its derivative, which are used to integrate the reduced dynamical system at a new parameter configuration. We provide two numerical experiments - the dynamical systems result from the semi-discretization of parametrized, nonlinear, hyperbolic partial differential equations that show, in addition to non-intrusivity, the proposed approach provides more stable and accurate approximations to each dynamical system across a large number of training and testing points than the popular empirical interpolation method. (C) 2020 Elsevier B.V. All rights reserved.

A Novel Efficient Method for Real-Time Computation of Parameterized Dynamic Equations with Large-Scale Dimension

A Novel Fast Dynamic Optimization Approach for Complex Multivariable Chemical Process Systems

An Efficient Technique for Recovering Responses of Parameterized Structural Dynamic Problems

Precise Time-Integration Method with Dimensional Expanding for Structural Dynamic Equations

An Efficient Method for Computation of Dynamic Responses for Large-Scale Structures Based on Time Integration and Galerkin Mapping

A model reduction method for parametric dynamical systems defined on complex geometries

A Robust Time-Integration Algorithm for Solving Nonlinear Dynamic Problems with Large Rotations and Displacements

An Energy Conservation Algorithm for Nonlinear Dynamic Equation

A tensor-train reduced basis solver for parameterized partial differential equations

Reduced-order Modelling of Parameter-Dependent, Linear and Nonlinear Dynamic Partial Differential Equation Models.

An efficient hybrid implicit time integration method with high accuracy for linear and nonlinear dynamics

A New Simple Method of Implicit Time Integration for Dynamic Problems of Engineering Structures

Non-intrusive model reduction of large-scale, nonlinear dynamical systems using deep learning

Reduced Krylov Basis Methods for Parametric Partial Differential Equations

Efficient method for computation of multi-parameterized problem based on least square mapping

PRECISE TIME-INTEGRATION WITH[1mm］ DIMENSION EXPANDING METHOD

An Adaptive Method Based on Local Dynamic Mode Decomposition for Parametric Dynamical Systems

Two-stage model reduction approaches for the efficient and certified solution of parametrized optimal control problems

Fast Solution for Large Scale Linear Algebraic Equations in Finite Element Analysis

Adaptive Dimensionality-Reduction for Time-Stepping in Differential and Partial Differential Equations

Dynamical model reduction method for solving parameter-dependent dynamical systems