Dr. Ziheng Chen, (ICMSEC, AMSS)

Numerical approximations of large deviations for invariant measures of parabolic SPDEs

Wednesday 25 November, 16:00-17:00 pm, N204

This talk considers the numerical approximations of large deviations for invariant measures of the ergodic semi-linear parabolic SPDEs with additive noise. To this end, we discrete the considered equation spatially by a spectral Galerkin method and temporally by an accelerated exponential Euler methods. In addition to verifying the ergodicity of these methods, we develop the large deviations principle for sample paths of spatial semi-discretization and the continued spatio-temporal full discretization, acting as an auxiliary results to establish the large deviations principle for invariant measures of both discretizations. The weakly asymptotically preserveing property between the numerical rate functions of the large deviations principles and the original ones are finally verified to confirm the effectiveness of our numerical approximations. 

欢迎大家参加！