Real-time control of a kite-model using an auto-generated nonlinear MPC algorithm

Authors

• H.J. Ferreau, B. Houska, K. Geebelen, M. Diehl

Reference

• In Proceedings of the
  18th IFAC World Congress,
  Milan, Italy, pages 2488 - 2493, September, 2011.

Abstract

Airborne wind energy systems like power-generating kites promise to become a sustainable and safe alternative for today’s fossil fuel based energy production. Controlling these systems is still a challenge due to the fast and highly nonlinear dynamics. A real-world prototype, which has been build at the K.U. Leuven, consists of a carousel which drives an airplane being attached at one of its arms. In the first part of this paper, we propose a nonlinear grey box model which is developed for controlling this kite carousel. In the second part, we present a code generation tool for nonlinear real-time MPC algorithms which exports plain C-code tailored to particular model dynamics. Numerical closed-loop simulations of the kite carousel show that auto-generated code allows to solve the resulting dynamic optimization problems within less than 900 microseconds.

Download

• Paper

Bibtex