Cooperative control with general linear dynamics and limited communication: Centralized and decentralized event-triggered control strategies

This paper presents event-triggered control techniques for the consensus problem with general linear dynamics. A novel consensus protocol is proposed, where each agent implements a model of the decoupled dynamics of its neighbors. We first provide a simple centralized condition to motivate the problem. Then, the focus is placed on designing decentralized consensus protocols. The decentralized approach proposed in this paper not only avoids the need for continuous communication between agents but also provides a decentralized and asynchronous method for transmission of information. This method gives more flexibility for scheduling information broadcasting compared to periodic and sampled-data implementations. Finally conditions are provided in order to guarantee positive inter-event times.