Asymptotic analysis of a certain random differential equation

We prove a limit theorem for the mathematical expectation of the solution of an initial value problem in a Hilbert space. The random differential equations considered here satisfy a strong mixing condition which is weaker than the one imposed in analogue results (Cogburn and Hersh, 1973; Papanicolaou and Varadhan, 1973). Our motivation to develop this analysis comes from a system (Nogueira, preprint) formed by coupling an external source to the Martin-Emch model (Martin and Emch, 1975).