Event-triggered Consensus for Multi-agent Systems with Asymmetric and Reducible Topologies

This paper studies the consensus problem of multi-agent systems with asymmetric and reducible topologies. Centralized event-triggered rules are provided so as to reduce the frequency of system's updating. The diffusion coupling feedbacks of each agent are based on the latest observations from its in-neighbors and the system's next observation time is triggered by a criterion based on all agents' information. The scenario of continuous monitoring is first considered, namely all agents' instantaneous states can be observed. It is proved that if the network topology has a spanning tree, then the centralized event-triggered coupling strategy can realize consensus for the multi-agent system. Then the results are extended to discontinuous monitoring, where the system computes its next triggering time in advance without having to observe all agents' states continuously. Examples with numerical simulation are provided to show the effectiveness of the theoretical results.