Expansion for the moments of a nonlinear stochastic model

We present a procedure to systematically evaluate all the moments of the Fokker-Planck equation by expanding them in a power series in a given function of t. The expansion coefficients are easily determined in terms of algebraic recursion relations. Applications to a linear Fokker-Planck equation, as well as to a truly nonlinear mean-field model, whose drift coefficient exhibits a functional dependence on the distribution function, show this formalism to be advantageous over the standard time series expansion of the moments which is shown to be rather impractical.Dirección General de Investigación Científica y Técnica (España) PB92-068