A Periodic Approach to Dynamic Output Feedback MPC for Quasi-LPV Model
Jianchen Hu, Baocang Ding
Abstract
This article considers dynamic output feedback model predictive control for quasi-linear parameter varying model with norm-bounded disturbance. The infinite horizon control moves are parameterized as the parameter-dependent dynamic output feedback (PDDOF) laws. In contrast to the traditional single-step approach where one PDDOF law is utilized, in this article a periodic approach where a sequence of PDDOF laws, which are invoked periodically in the future predictions, are applied. The regions of attraction for the periodic approach are considerably larger, and the control performance improved, as compared with the single-step approach. The recursive feasibility, and convergence of augmented state and output/input, are guaranteed. An illustrative example is given to show the effectiveness of this approach.