Event-triggered control systems under packet losses

Networked control systems (NCSs) offer many benefits in terms of increased flexibility and maintainability but might also suffer from inevitable imperfections such as packet dropouts and limited communications resources. In this paper, (static and dynamic) event-triggered control (ETC) strategies are proposed that aim at reducing the utilization of communication resources while guaranteeing desired stability and performance criteria and a strictly positive lower bound on the inter-event times despite the presence of packet losses. For the packet losses, we consider both configurations with an acknowledgement scheme (as, e.g., in the transmission control protocol (TCP)) and without an acknowledgement scheme (as, e.g., in the user diagram protocol (UDP)). The proposed design methodology will be illustrated by means of a numerical example which reveals tradeoffs between the maximum allowable number of successive packet dropouts, (minimum and average) inter-event times and Lp-gains of the closed-loop NCS.