Spectra, exit times and long time asymptotics in the zero-white-noise limit
- 1 November 1995
- journal article
- research article
- Published by Taylor & Francis in Stochastics and Stochastic Reports
- Vol. 55 (1-2), 1-20
- https://doi.org/10.1080/17442509508834015
Abstract
For ϵ>0, we consider the valued Markov process whose generator is . X ϵ is a small random perturbation of the deterministic dynamical system . We first establish rough asymptotics for the eigenvalues of the operator with Dirichlet of Neumann boundary conditions when ϵ tends to 0. Using these estimates, we describe the asymptotics of where is the exit time of X ϵ out of a fixed bounded domain. We are also able to express the limit law of X ϵ at time exp as a linear combination of the equilibrium measure of X ϵ restricted to some domains of . We derive a more precise description of X ϵ giving sufficient conditions for the unpredictability of and metastability. Our approach is based on spectral estimates and ultra-contractivityKeywords
This publication has 7 references indexed in Scilit:
- Limit theorems for diffusions with a random potentialStochastic Processes and their Applications, 1995
- Zero White Noise Limit through Dirichlet forms, with application to diffusions in a random mediumProbability Theory and Related Fields, 1994
- Small random perturbations of finite- and infinite-dimensional dynamical systems: Unpredictability of exit timesJournal of Statistical Physics, 1989
- Asymptotics of the spectral gap with applications to the theory of simulated annealingJournal of Functional Analysis, 1989
- Metastability for a Class of Dynamical Systems Subject to Small Random PerturbationsThe Annals of Probability, 1987
- Random Perturbations of Dynamical SystemsPublished by Springer Nature ,1984
- On the exponential exit law in the small parameter exit problemStochastics, 1983