Lower estimates of transition densities and bounds on exponential ergodicity for stochastic PDEs (Q850975)

The paper is devoted to ergodic properties of solutions to a semilinear stochastic equation \[ dX^x=(AX^x+F(X^x))dt+\sqrt{Q}dW \] considered in separable Banach space \(E\), where \(W\) is a cylindrical Wiener process on a Hilbert space \(H, E\subset H\). A formula for transition density of this process is given in terms of the Ornstein-Uhlenbeck bridge and a lower estimate of the density is provided. There are provided computable bounds on the convergence rates and the spectral gap for the Markov semigroups defined by these equations. Examples are also given.