Offline robust MPC based on polyhedral invariant set

An offline method of synthesizing constrained robust model predictive controller based on polyhedral invariant set is proposed. This algorithm offline constructs a sequence of invariant sets and local state feedback controllers which can keep the states inside the invariant set even the uncertainty exists. Online it introduces perturbations which increase extra degrees of freedom to give better performance. It also adds terminal constraint set which drives the states into the smaller invariant set to achieve robust stability. The online and offline computation of the algorithm can be solved easily with the tool YALMIP of MATLAB. This algorithm dramatically reduces the online computation, guarantees the control performance by adding the external signal input and decreases the conservation with polyhedral invariant set instead of ellipsoidal invariant set.