Observer-based networked predictive controller with bounded and unknown time-varying delay

Abstract

This paper investigates a networked control system with bounded and time-varying delays. Also, an observer-based predictive controller is developed for the active compensation of the network communication delay since it may lead to poor performances or even unstable dynamics for the systems. The existence of the state observer can be established by choosing an appropriate Lyapunov function, and to do this, Luenberger observer gain is computed with Linear Matrix Inequality (LMI) based on the stability conditions of the system. The practical aspect of this research, different from previous works, is the accessibility of the states that are not always available for the system. Controller and observer gains and other essential variables are derived through LMI. In the closed-loop system, which is modeled as a switching system, the switches are based on communication delays. Using the Lyapunov stability method for switching systems, sufficient LMI conditions are derived to guarantee the stability of the closed-loop system. Finally, the results of the simulation have demonstrated the performance of the methodology presented.