On a global optimization algorithm for bivariate smooth functions (Q481774)

The authors consider the problem of minimizing a twice continuously differentiable function \(f(x)\) of two variables over a compact two-dimensional set \(A\). The aim is to approximate the global minimum value of the objective function under the given constraint. A method based on statistical models using a partition of the feasible region is proposed. The partition is carried out by the Delaunay triangulation. The asymptotic convergence rate of the proposed method is studied. Possibilities of further extensions and generalizations of the method are discussed. Numerical examples illustrating the theoretical results are presented.