Time-domain Smith predictor in discrete-time networked control systems

Abstract

Translate Article

In the paper, we study the stability of the discrete-time networked control systems. If the network link is with a limited communication capacity, the network-induced delay is unavoidable. Considering the network-induced delay, the closed-loop networked control systems are usually subject to the delayed state. The traditional method for the study of the networked control systems is mainly to construct a Lyapunov function and derive the stability criterion. Generally, the derived stability condition is only sufficient. The selection of the Lyapunov function is attracting increasing attention as a better Lyapunov function may lead to less conservative results. In this work, we use an alternate way to get the less conservative results. A time-domain Smith predictor is employed to predict the non-delayed state. Then, the control law is the feedback of the delayed measurements and the non-delayed prediction. Simulation results are given to show that the proposed control law can expand the upper bounds with the existing relative conservative stability criterion.