Predictive control sensitivity (Q1289415)

The receding horizon continuous-time predictive controller is considered. It is assumed that both, the modelled and the real process are controllable and the optimization problem has a unique solution. It is shown that the controller gain increases to infinity when the current control horizon reaches zero. As a consequence the receding horizon predictive controller realizes the target state in the presence of additive process disturbances. The required control signal may take significantly large values but remains bounded even at the final part of the horizon. In case of saturations the target state can be reached only asymptotically.