Stochastic generalized Burgers equations driven by fractional noises (Q652510): Difference between revisions

This paper studies the existence and uniqueness of solutions for a class of stochastic generalized Burgers equations driven by additive multi-parameter fractional noise of the form \[ \frac{\partial u}{\partial t}=\Delta u + \frac{\partial g}{\partial x}(u) + \dot{W}^{H} \] in one spatial dimension with Dirichlet boundary conditions. The function \(g\) is a polynomial of order \(3\) with positive leading coefficient and \(\dot{W}^{H}\) is the derivative of the two-parameter fractional Brownian motion (understood in the distributional sense). The range of Hurst parameters used is \(H=(H_{1},H_{2})\), \(H_{i} \in (1/2,1)\), \(i=1,2\). The treatment of the equation is done by rewriting it in mild form in terms of a nonlinear integral equation employing the Green function \(G\) of the operator \(\frac{\partial}{\partial t}-\Delta\) with Dirichlet boundary conditions \[ u(t,x)=\int_{0}^{1}G(t,x,y)u_{0}(y)\,dt - \int_{0}^{t}\int_{0}^{} \frac{\partial}{\partial y} G(t-s,x,y) g(u(s,y)\,dy ds + L(t,x), \] where \(L(\cdot, \cdot)\) is the stochastic convolution defined by \[ L(t,x):=\int_{0}^{t}\int_{0}^{1}G(t-s,x,y) W^{H}(dy,ds). \] Using detailed estimates of the stochastic convolution, and a truncation argument the authors prove the existence of a unique adapted solution \(u \in L^{\infty}([0,T]; L^{p}([0,1]))\) (\(p \geq 3\)) with a.s. continuous trajectories. Furthermore, using techniques from the Malliavin calculus for fractional processes and a suitable approximation scheme, the authors show that the law generated by the Burgers equation driven by additive fractional noise is absolutely continuous with respect to the Lebesgue measure.