Synthesis of off-line robust MPC based on polyhedral invariant set

The online computation efficiency is always an important issue of the robust model predictive control (RMPC), especially for fast systems. For uncertain systems with polytopic description, two synthesis approaches of off-line RMPC are presented. By offline designing several polyhedral invariant sets properly, the online computation of RMPC is only simple algebra calculation, which reduces the online computational burden greatly. Polyhedral invariant set enlarges the feasible region of RMPC compared with the previous designs based on ellipsoidal invariant set. The closed-loop stability and recursive feasibility can be guaranteed.