Distributed Dynamic Event-Triggered Communication Mechanisms for Dynamic Average Consensus

This paper studies the dynamic average consensus problem of multi-agent systems under event-triggered communication. In this problem, each agent has access to a time-varying reference signal and aims to track the average of all reference signals. Distributed algorithms with event-triggered communication have been developed to achieve dynamic average consensus. Nevertheless, these existing event-triggered communication mechanisms cannot guarantee the existence of a designable positive minimum inter-event time (MIET), which is important in their practical implementation. Motivated by this observation, we propose a distributed dynamic event-triggered communication mechanism (ETCM) for each agent. It is shown that the proposed ETCM guarantees the existence of a positive MIET that is locally adjustable by tuning design parameters. It is also shown that the dynamic average consensus is achieved with any pre-specified level of accuracy. As an illustrative example, the theoretical results are applied to a networked battery energy storage system for state-of-charge balancing and desired total power tracking.