Branch-and-lift algorithm for deterministic global optimization in nonlinear optimal control (Q467420)

The paper considers nonlinear optimal control problems. The authors develop a global optimal control algorithm based on a generic spatial branch-and-bound algorithm. A new lifting operation is introduced which refines the control parametrization via a Gram-Schmidt orthogonalization process, while at the same time eliminating control subregions that are either infeasible or cannot contain any global optima. They demonstrate the applicability of their new algorithm on a numerical case study.