A Condensing Algorithm for Nonlinear MPC with a Quadratic Runtime in Horizon Length

  • Joel A E Andersson
  • Moritz Diehl
  • Janick V Frasch
  • Milan Vukov
    • Categories Optimization of Simulated Systems, Quadratic Programming Tags , , , , Short URL: https://optimization-online.org/?p=14384

    A large number of practical algorithms for Optimal Control Problems (OCP) relies on a so-called condensing procedure to exploit the given structure in the quadratic programming (QP) subproblems. While the established structure-exploiting condensing algorithm is of cubic complexity in the horizon length, in this technical note we propose a novel algorithm that is only of quadratic complexity in horizon length. We present numerical results confirming that the proposed algorithm is faster in practice and discuss implications for related QP solution algorithms that rely on the elimination of the state sequence.

    Citation

    Electrical Engineering Department (ESAT) and Optimization in Engineering Center (OPTEC), KU Leuven, Belgium (2013)

    Article

    Download

    View A Condensing Algorithm for Nonlinear MPC with a Quadratic Runtime in Horizon Length