Data-driven dynamic periodic event-triggered control for unknown linear systems: A hybrid system approach

Abstract

This paper investigates the problem of data-driven dynamic periodic event-triggered mechanism (DPETC) for unknown linear time-invariant (LTI) systems. First, we construct a set of all the systems that are compatible with the collected data. In order to stabilize the true system, we need to stabilize all the systems in the set. Next, we introduce the dynamic periodic event-triggered mechanism and the entire system is reformulated as a hybrid system. Sufficient conditions for $L_2$-stability of the hybrid system are derived under some assumptions. Then we present a new approach to obtain data-dependent synthesis conditions in the form of linear matrix inequalitys (LMIs). Finally, a numerical example and a hardware-in-the-loop (HIL) example are given to verify the effectiveness of our approach.