On a Globalized Augmented Lagrangian-SQP Algorithm for Nonlinear Optimal Control Problems with Box Constraints

A globalized augmented Lagrangian-SQP algorithm for nonlinear optimal control problems with box constraints on the control is presented. For the solution of the quadratic subproblems of the SQP-type algorithm a very efficient primal-dual method yielding exact subproblem solutions is utilized. The globalization strategy is based on a suitable modification of the Hessian of the Lagrangian and on a line search utilizing an exact penalty functional. Finally, a report on numerical test runs is given.