Model predictive control for NCS with one sampling transmission delay

This paper proposes a finite switching horizon model predictive control for networked control system with less than one sampling state transmission delay. The uncertain state only has two cases: state delay occurs or not. In the first case, the controller input is the former state of the system. In the second case, the controller input is the current state of the system. The control strategy which addresses both the two cases is characterized as a constrained optimization problem of the quadratic cost over a finite switching horizon at each sampling instant. A receding horizon implementation of this strategy in general renders satisfaction of the desirable input and state constraints. For the implementation of the controller, a linear matrix inequality approach is developed. Also, the stability analysis is addressed. A numerical example is given to illustrate the feasibility and effectiveness of the proposed algorithm.