Two-dimensional global optimization based upon a continuous Wiener model

A method of global optimization is presented which extends Kushner´s one-dimensional method into higher dimensions. This technique differs from other n-dimensional extensions since it retains the properties of Kushner´s original method, models the unknown function by a stochastic process which is continuous over the search space, and is capable of finding solutions along the boundaries of the space. This formulation of the problem allows an analytic solution for the point of maximum probability for the entire search space which explicitly considers the boundaries of the search space as possible locations for the global solution