Rough Path Theory To Approximate Random Dynamical Systems
Gao Hongjun,Garrido-Atienza María J.,Gu Anhui,Lu Kening,Schmalfuss Björn,María J. Garrido-Atienza,Björn Schmalfuß
DOI: https://doi.org/10.1137/20M1325022
2021-01-01
SIAM Journal on Applied Dynamical Systems
Abstract:We consider the rough differential equation dY = f (Y)d omega, where omega = (omega, omega) is a rough path defined by a Brownian motion w on R-m Under the usual regularity assumption on f, namely, f is an element of C-b(3)(R-d, R-d x m) the rough differential equation has a unique solution that defines a random dynamical system phi(0). On the other hand, we also consider an ordinary random differential equation dY(delta) = f (Y-delta)d omega(delta), where omega(delta) is a random process with stationary increments and continuously differentiable paths that approximates omega. The latter differential equation generates a random dynamical system phi delta as well. We show the convergence of the random dynamical system phi delta to phi(0) for delta -> 0 in Holder norm.