Convergence and application of a decomposition method using duality bounds for nonconvex global optimization (Q700712): Difference between revisions

A decomposition method is studied for solving a class of global optimization problems of the form: Minimize \(F(x, y)\) subject to \((x, y)\in Z\), where \[ Z\equiv\{(x, y)\mid G_i(x, y)\leq 0,\;i=1,\dots,m, x\in X, y\in Y\}, \] \(X\), \(Y\) are closed subsets of \(\mathbb{R}^n\) and \(\mathbb{R}^p\), respectively, \(F\) and \(G_i\) are continuous functions defined on a set containing \(X\times Y\). The variables of this problems are divided in two groups such that in each group, the functions involved have the same structure ( e.g. they are linear, convex, concave etc.). The problem is solved using the decomposition idea suggested by Benders. To solve the resulting master problem, which is in fact the optimal value function of a nonlinear parametric optimization problem, a branch-and-bound scheme is proposed. The branching procedure is performed by using some techniques of polyhedral subdivisions from global optimization and the estimation of the lower bounds is based on the weak duality theorem in Lagrange duality.