Averaging principle for stochastic variational inequalities with application to PDEs with nonlinear Neumann conditions (Q2139620)

Stochastic variational inequalities have been widely used in various areas including optimizations and game theory subject to uncertainties, where some data are described by functions which may not be smooth or even continuous. In this paper, the authors consider a stochastic dynamic system in terms of multiscale stochastic variational inequalities. They establish averaging principles for a separated time-scale system of fully coupled stochastic system characterized by stochastic variational inequalities. Under non-Lipschitz continuous conditions, they show that the classical weak convergence result holds for this type of stochastic systems. Strong convergence is also studied for the cases when the diffusion coefficients of the slow motions do not depend on the fast motion components. As an application of the averaging principle, they study the homogenization of generalized backward SDEs and Neumann-type semilinear parabolic variational inequalities.