A self-dual variational approach to stochastic partial differential equations (Q2422463): Difference between revisions

In this paper, the authors show how self-dual variational calculus leads to variational solutions of various stochastic partial differential equations driven by monotone vector fields. It is considered the stochastic partial differential equations of the following form \(du(t)=-A(t,u(t))dt-\Lambda u(t)dt+B(t,u(t))dW(t)\), \(u(0)=u_0\), where \(u_0\in L^2(\Omega,\mathcal{F}_0,P;H)\), \(H\) being a Hilbert space, and \(W(t)\) is a real-valued Wiener process, the vector field \(A:\Omega\times[0,T]\times V\to V^{*}\) is an adapted random maximal monotone map, where \(V\) is a Banach space such that \(V\subset H\subset V^{*}\) constitute a Gelfand triple. The operators \(\Lambda: V\to V^{*}\) and \(B: [0,T]\times V\to H\) are the suitable maps. The authors show how one can lift self-dual Lagrangians from state space to function spaces and then construct random self-dual Lagrangians on Itô spaces of stochastic processes. It is presented a variational resolution for basic stochastic partial differential equations involving additive noise by establishing a stochastic version of the well-known Bolza duality. The authors study the general stochastic partial differential equations driven by a self-dual Lagrangian on \(L^{\alpha}(\Omega_ T;V)\times L^{\beta}(\Omega_ T;V^{*})\) and a non-additive noise. Then authors apply this result to resolve equations of the form \(d u(t)=\Delta udt+\vert u\vert^{q-1} udW(t)\) in \([0,T]\times D\), \(u(t,x)=0\) in \([0,T]\times\partial D\), \(u(0,x)=u_0(x)\) on \(D\), where \(1/2\le q<n/(n-2)\), \(D\) is a bounded domain in \(\mathbb R^ n\). The stochastic Navier-Stokes equations in two dimensional case is considered.