Robust receding-horizon control for linear systems with model uncertainties

A method for designing a robust receding-horizon controller for linear systems with model uncertainties is proposed. An upper bound of the worst-case finite-horizon performance index is calculated and then a state-feedback gain, which minimizes the upper bound, is obtained. Using Lyapunov arguments, it is shown that the feasible control law asymptotically stabilizes the closed-loop system. This can be achieved by imposing a terminal inequality condition at the terminal time of the finite-horizon performance index. This design method for the robust receding-horizon controller is extended to tracking problems. Finally some numerical examples are suggested in order to show how these design methods work in real plants compared with the conventional receding-horizon controller