Consensus of multi-agent systems with time-varying topology: An event-based dynamic feedback scheme

The event-based control strategy is an effective methodology for reducing the controller update and communication over the network. In this paper, the event-based consensus of multi-agent systems with linear dynamics and time-varying topology is studied. For each agent, a state-dependent threshold with an exponentially decaying bound is presented to determine the event times, and a new event-based dynamic feedback scheme is proposed. It is shown that the controller update for each agent is only dependent on its own event times, which reduces significantly the controller update or computation for each agent. Moreover, based on the event-based dynamic feedback scheme and the event triggering function presented in this paper, the continuous communication among neighboring agents is avoided, and the Zeno-behavior of the closed-loop systems is excluded. A numerical example is given to illustrate the effectiveness of theoretical results. Copyright (c) 2016 John Wiley & Sons, Ltd.