Interaction between control and estimation in nonlinear MPC

Authors:
  • MHMorten Hovd
    Morten Hovd
  • RBRobert R. Bitmead
    Robert R. Bitmead

Abstract
This paper shows how to take the quality of the state estimation into account in the formulation of the optimization criterion for model predictive control (MPC). This is relevant for the control of nonlinear plants, for which the separation principle in general does not apply. The method is illustrated on an example which is locally weakly unobservable at the reference state