Large scale direct optimal control applied to the re-entry problem
Résumé
We present the numerical solution of an atmospheric reentry problem for a space shuttle. We discretize the control and state with the same grid, and use a large-scale successive quadratic programming technique. With the help of sliding horizon and successive refinement of the discretization, we can solve on a workstation a problem with 1600 time step, an unusually large figure for this kind of real world optimal control problem.