Averaging principle for semilinear slow-fast rough partial differential equations (2411.06089v2)

Abstract: In this paper, we investigate the averaging principle for a class of semilinear slow-fast partial differential equations driven by finite-dimensional rough multiplicative noise. Specifically, the slow component is driven by a general random $\gamma$-H\"{o}lder rough path for some $\gamma \in (1/3,1/2)$, while the fast component is driven by a Brownian rough path. Using controlled rough path theory and the classical Khasminskii's time discretization scheme, we demonstrate that the slow component converges strongly to the solution of the corresponding averaged equation under the H\"{o}lder topology.