Command Governor Based Adaptive Control for Constrained Linear Systems

This paper presents a novel control method for constrained linear systems with model uncertainties. The proposed method achieves a near performance to the reference system, adaptability to the model uncertainties and meet all the state-constraints at the same time. The design of such a method involves a command governor and a model reference adaptive controller. The command governor adjusts the command signal based on the information of the nominal system to make the nominal system behaviors meet all the constraints. The adaptive controller compensates the model uncertainties of the actual plant, which makes the system to mimic the nominal closed-loop system. As a consequence, the real system can realize a quite near performance of the nominal system with stability and robustness, without violating any constraints.