Enhanced Robust NMPC Based on Nominal Predictions

Abstract This paper deals with the design of robust predictive controllers for uncertain constrained nonlinear systems. The considered controller is based on nominal predictions and can be posed as a standard MPC. Robust feasibility is ensured by imposing more conservative constraints in the optimization problem to take into account explicitly the effect of the uncertainty. This is done by means of the calculation of a sequence of sets, which can be easily obtained exploiting the uniform continuity of the model function. Sufficient conditions to prove input-to-state stability of the proposed controller are presented. The proposed approach can be considered an enhanced and updated formulation of a previous approach presented by the authors aimed to the reduction of the conservativeness.