English

Some strong limit theorems in averaging

Probability 2024-06-21 v4

Abstract

The paper deals with the fast-slow motions setups in the discrete time Xϵ((n+1)ϵ)=Xϵ(nϵ)+ϵB(Xϵ(nϵ),ξ(n))X^\epsilon((n+1)\epsilon)=X^\epsilon(n\epsilon)+\epsilon B(X^\epsilon(n\epsilon),\xi(n)), n=0,1,...,[T/ϵ]n=0,1,...,[T/\epsilon] and the continuous time dXϵ(t)dt=B(Xϵ(t),ξ(t/ϵ)).t[0,T]\frac {dX^\epsilon(t)}{dt}=B(X^\epsilon(t),\xi(t/\epsilon)).\, t\in [0,T] where BB is a smooth vector function and ξ\xi is a sufficiently fast mixing stationary stochastic process. It is known since 1966 (Khasminskii) that if Xˉ\bar X is the averaged motion then Gϵ=ϵ1/2(XϵXˉ)G^\epsilon=\epsilon^{-1/2}(X^\epsilon-\bar X) weakly converges to a Gaussian process GG. We will show that for each ϵ\epsilon the processes ξ\xi and GG can be redefined on a sufficiently rich probability space without changing their distributions so that Esup0tTGϵ(t)G(t)2M=O(ϵδ)E\sup_{0\leq t\leq T}|G^\epsilon(t)-G(t)|^{2M} =O(\epsilon^{\delta}), δ>0\delta>0 which gives also O(ϵδ/3)O(\epsilon^{\delta/3}) Prokhorov distance estimate between the distributions of GϵG^\epsilon and GG. In the product case B(x,ξ)=Σ(x)ξB(x,\xi)=\Sigma(x)\xi we obtain almost sure convergence estimates of the form sup0tTGϵ(t)G(t)=O(ϵδ)\sup_{0\leq t\leq T}|G^\epsilon(t)-G(t)|=O(\epsilon^\delta) a.s., as well as the functional form of the law of iterated logarithm for GϵG^\epsilon. We note that our mixing assumptions are adapted to fast motions generated by important classes of dynamical systems.

Keywords

Cite

@article{arxiv.2209.10364,
  title  = {Some strong limit theorems in averaging},
  author = {Yuri Kifer},
  journal= {arXiv preprint arXiv:2209.10364},
  year   = {2024}
}

Comments

arXiv admin note: text overlap with arXiv:2105.01940, arXiv:2111.05390

R2 v1 2026-06-28T01:49:14.265Z