Computer algebra derives correct initial conditions for low-dimensional dynamical models

DOI10.1016/S0010-4655(99)00494-4MaRDI QIDQ1568593zbMATH OpenOpenAlexFDO

Authors A. J. Roberts

Publication date 21 June 2000

Published in Computer Physics Communications (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/chao-dyn/9901010

zbMATH Keywords

centre manifoldscomputer algebracorrect initial conditionsdual equationdynamical systemseigenvaluesiterative methodlow-dimensional models

Mathematics Subject Classification ID

Dynamics induced by flows and semiflows (37C10) Dynamical systems involving maps of the interval (37E05) Discretization methods and integrators (symplectic, variational, geometric, etc.) for dynamical systems (37M15) Numerical methods for Hamiltonian systems including symplectic integrators (65P10) Symbolic computation and algebraic computation (68W30)

Abstract: To ease analysis and simulation we make low-dimensional models of complicated dynamical systems. Centre manifold theory provides a systematic basis for the reduction of dimensionality from some detailed dynamical prescription down to a relatively simple model. An initial condition for the detailed dynamics also has to be projected onto the low-dimensional model, but has received scant attention. Herein, based upon the reduction algorithm in~cite{Roberts96a}, I develop a straightforward algorithm for the computer algebra derivation of this projection. The method is systematic and is based upon the geometric picture underlying centre manifold theory. The method is applied to examples of a pitchfork and a Hopf bifurcation. There is a close relationship between this projection of initial conditions and the correct projection of forcing onto a model. I reaffirm this connection and show how the effects of forcing, both interior and from the boundary, should be properly included in a dynamical model.

Review text

This paper is based on earlier work by the author who gave an iterative method for the construction of low-dimensional models from more complicated dynamic models by the elimination of uninteresting fast modes of behaviour. In this present paper, computer algebra is used to derive initial conditions for these low-dimensional models in order that accurate forecasting of the dynamics can be made. The cases of centre manifolds formed when a set of eigenvalues is zero or have non-zero imaginary part (substantially more difficult) are considered. This is done by projecting the initial conditions onto the centre manifold along the isochronic manifold, which is achieved by computing the space of normals to the isochronic manifold through iteratively refining an approximation based upon a residual process for a projected dual equation.

Recommendations

• Low-dimensional modelling of dynamics via computer algebra
• Initial conditions for models of dynamical systems
• scientific article; zbMATH DE number 2145063
• Computer algebra methods in the study of nonlinear differential systems
• Holistic projection of initial conditions onto a finite difference approximation

Cited in

(13)

• Low-dimensional modelling of dynamics via computer algebra
• Initial conditions for models of dynamical systems
• Global linearization and fiber bundle structure of invariant manifolds
• A holistic finite difference approach models linear dynamics consistently
• Computation of bifurcation margins based on robust control concepts
• Computational singular perturbation method for nonstandard slow-fast systems
• Spectral analysis of the Koopman operator for partial differential equations
• A unified approach to compute foliations, inertial manifolds, and tracking solutions
• Holistic discretization ensures fidelity to Burgers' equation
• Holistic projection of initial conditions onto a finite difference approximation
• Embed in ensemble to rigorously and accurately homogenize quasi-periodic multi-scale heterogeneous material
• Accurate families of multi-continuum micromorphic homogenisations in multi-D space-time via dynamical systems theory
• General oblique projections for model reduction via spectral submanifolds

This page was built for publication: Computer algebra derives correct initial conditions for low-dimensional dynamical models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1568593)