Biblio

This paper studies a distributed event-triggered communication and control strategy that solves the multi-agent average consensus problem. The proposed strategy does not rely on the continuous or periodic availability of information to an agent about the state of its neighbors, but instead prescribes isolated event times where both communication and controller updates occur. In addition, all parameters required for its implementation can be locally determined by the agents. We show that the resulting network executions are guaranteed to converge to the average of the initial agents' states, establish that events cannot be triggered an infinite number of times in any finite time period (i.e., absence of Zeno behavior), and characterize the exponential rate of convergence. We also provide sufficient conditions for convergence in scenarios with time-varying communication topologies. Simulations illustrate our results.