The parameterization method in dynamical systems

DOI10.2436/20.2002.01.40MaRDI QIDQ2906532zbMATH OpenFDO

Publication date 5 September 2012

Published in Butlletí de la Societat Catalana de Matemàtiques (Search for Journal in Brave)

Full work available at URL https://doi.org/10.2436/20.2002.01.40

invariant manifolds KAM theory parametrization methods small divisors

Invariant manifolds for ordinary differential equations (34C45) Asymptotic properties of solutions to ordinary differential equations (34D05) Topological and differentiable equivalence, conjugacy, moduli, classification of dynamical systems (37C15) Generic properties, structural stability of dynamical systems (37C20) Perturbations of finite-dimensional Hamiltonian systems, normal forms, small divisors, KAM theory, Arnol'd diffusion (37J40)

Review text

The present work, written in Catalan, is a survey on the parametrization method in dynamical systems, providing an unified description of its applications to prove the existence and explicitly compute some invariant objects in a dynamical system, either discrete or continuous.NEWLINENEWLINEAssume, to simplify the exposition, that \(F:\mathcal{M}\to \mathcal{M}\) is a dynamical system, that is, a smooth diffeomorphism on a manifold. The ``parametrization method consists in finding solutions \(P\) and \(f\) of the equation NEWLINE\[NEWLINE F\circ P = P \circ f, NEWLINE\]NEWLINE the ``invariance equation, where \(P:N \to \mathcal{M}\) is an embedding and \(f:N\to N\) is a diffeomorphism. With this notation, \(P(N)\) is an invariant manifold by \(F\), while \(f\) describes the dynamics of~\(F\) restricted to~\(P(N)\).NEWLINENEWLINEThe parametrization method, as described in this paper, has two different steps. The first one consists in finding approximate solutions of the invariance equation by solving a set of homological equations. This part depends strongly on the object under consideration, and is described with certain detail here. The second one provides a true solution of the equation by rewriting the invariance equation as a suitable fixed point equation, which can be solved by different arguments if the formal part can be obtained up to high enough order. This second step if naturally omitted in this exposition.NEWLINENEWLINEThe present work describes how to solve the invariance equation in three different contexts.NEWLINENEWLINEThe first one, which is the simplest, deals with the stable or unstable manifolds of fixed points of the map~\(F\). In this case, the graph transform method is a particular instance of the parametrization method, although not necessarily the most advantageous in many situations.NEWLINENEWLINEThe second one describes how to apply the method to find the continuation of a normally hyperbolic cylinder.NEWLINENEWLINEThe last part focuses on the existence of invariant tori with quasiperiodic dynamics, in the case where the map~\(F\) is exact symplectic. It should be remarked that in this last case there is no need neither of action-angle variables (to be close to an integrable map close to the invariant torus) nor of a canonical two-form. In this way, one can use this method as an a posteriori argument: finding an approximate solution of the invariance equation ensures the existence of the a true solution. Consequently, one can use this method to implement computed assisted proofs. Note that in this case it is necessary to use KAM arguments in order to find the true solution of the invariant equation since small divisors appear in solving the homological equations.

Recommendations

Cited in

(7)

This page was built for publication: The parameterization method in dynamical systems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2906532)