Design and convergence analysis of numerical methods for stochastic evolution equations with Leray–Lions operator

Jérôme Droniou, Beniamin Goldys and Kim-Ngan Le


The gradient discretisation method (GDM) is a generic framework, covering many classical methods (Finite Elements, Finite Volumes, Discontinuous Galerkin, etc.), for designing and analysing numerical schemes for diffusion models. In this paper, we study the GDM for a general stochastic evolution problem based on a Leray–Lions type operator. The problem contains the stochastic \(p\)-Laplace equation as a particular case. The convergence of the Gradient Scheme (GS) solutions is proved by using Discrete Functional Analysis techniques, Skorohod theorem and the Kolmogorov test. In particular, we provide an independent proof of the existence of weak martingale solutions for the problem. In this way, we lay foundations and provide techniques for proving convergence of the GS approximating stochastic partial differential equations.

Keywords: \(p\)-Laplace equation, stochastic PDE, numerical methods, gradient discretisation method, convergence analysis.

AMS Subject Classification: Primary 65M08; secondary % FV methods %65M12, 65M25, 65M60, 76S05.

This paper is available as a pdf (424kB) file.

Monday, May 11, 2020