Improved model prediction and RMPC design for LPV systems with bounded parameter changes

This paper studies the future model prediction and robust model predictive control (RMPC) design for linear parameter varying systems with bounded parameter changes. By developing tight bound estimations for varying parameters, we construct a set-valued map as the predicted family of future models. This construction attains accurate estimations and thus reduces conservativeness. Based on model predictions, we use a parameter-dependent feedback to design RMPC that achieves an enhanced performance with guaranteed robust and stability properties.