Systems with persistent disturbances: Predictive control with restricted constraints (Q5939327)

A robust predictive control strategy of linear discrete-time systems subject to simultaneous hard state and control constraints and persistent bounded disturbances is proposed. It is shown that the insertion of suitably chosen constraint restrictions robustifies the predictive controller with respect to the bounded disturbances and ensures feasibility for all admissible disturbance realizations, provided the initial state is feasible. The insertion of constraint restrictions improves the control strategy in several directions: (i) the multivariable systems and general state constraints can be treated, (ii) any known stabilizing terminal control law can be used in the predictive control algorithm, (iii) the proposed robust predictive controller guarantees asymptotic stability in the disturbance-free case and boundedness of the state response in the presence of disturbances. Moreover, in the latter case, convergence of the state to the smallest robust invariant set is ensured. Several examples are chosen to demonstrate the effectivenss of the proposed strategy.