Exponential mixing properties of stochastic PDEs through asymptotic coupling (Q1864421): Difference between revisions

The author considers parabolic stochastic partial differential equations driven by white noise in time. More precisely, the existence, uniqueness, and speed of convergence towards the invariant measure for such systems is investigated. He considers the following PDE \[ dx=Axdt+F(x)dt+Qd\omega (t),\quad x(0)=0,\tag{1} \] where \(x\) belongs so some Hilbert space \({H}\), \(A\) is the generator of a \(C_{0}\)-semigroup on \({H}\), \(F:{H}\) is some nonlinearity, \(\omega\) is the cylindrical Wiener process on some other space \({W}\) and \(Q:{W} \rightarrow {H}\) is a bounded operator. Then he proves that the solution of the equation (1) is asymptotically stable if there exists a unique probability measure \(\mu _{\ast }\) on \({H}\) such that the laws of \(x(t)\) converge to \(\mu _{\ast }\), independent of the initial condition \(x_{0}.\) Finally several examples are investigated, including someones where the noise does not act on every determining mode directly.