Konvergenzuntersuchnng zu einem verfahren der zulässige-n bichtungen bei nichtlinearen optimierungsaufgaben

Abstract

For a nonlinear programming problem with convex and differentiable restriction functions and an objective function, the gradient of which satisfies a Lipschitz cindition, a solution procedure is presented. For this prodedure convergence to a stationary point is proved. For the case, in which the objective function is strictly and boundedly convex, estimations on the convergence behaviour are or obtained.