15863 articles – 31997 Notices  [english version]
HAL : hal-00694470, version 1

Fiche concise  Récupérer au format
Stochastic CGL equations without linear dispersion in any space dimension
Kuksin S., Nersesyan V.
http://hal.archives-ouvertes.fr/hal-00694470
Preprint, Working Paper, Document sans référence, etc.
Mathématiques/Equations aux dérivées partielles
Stochastic CGL equations without linear dispersion in any space dimension
Sergei Kuksin () 1, Vahagn Nersesyan () 2
1 :  Centre de Mathématiques Laurent Schwartz (CMLS-EcolePolytechnique)
http://www.math.polytechnique.fr
CNRS : UMR7640 – Polytechnique - X
Ecole Polytechnique 91128 Palaiseau
France
2 :  Laboratoire de Mathématiques de Versailles (LM-Versailles)
http://www.math.uvsq.fr
CNRS : UMR8100 – Université de Versailles Saint-Quentin-en-Yvelines
45, avenue des Etats-Unis 78035 VERSAILLES cedex
France
We consider the stochastic CGL equation $$ \dot u- \nu\Delta u+(i+a) |u|^2u =\eta(t,x),\;\;\; \text {dim} \,x=n, $$ where $\nu>0$ and $a\ge 0$, in a cube (or in a smooth bounded domain) with Dirichlet boundary condition. The force $\eta$ is white in time, regular in $x$ and non-degenerate. We study this equation in the space of continuous complex functions $u(x)$, and prove that for any $n$ it defines there a unique mixing Markov process. So for a large class of functionals $f(u(\cdot))$ and for any solution $u(t,x)$, the averaged observable $\E f(u(t,\cdot))$ converges to a quantity, independent from the initial data $u(0,x)$, and equal to the integral of $f(u)$ against the unique stationary measure of the equation.
Anglais

Liste des fichiers attachés à ce document : 
PDF
Paper.pdf(294.9 KB)
PS
Paper.ps(914.6 KB)