Dynamic event-triggered control of heterogeneous multi-agent systems based on multi-order neighbors’ information

Abstract

In this paper, the consensus problems of discrete-time heterogeneous multi-agent systems are investigated under a directed topology, in which the dynamics of agents are first-order and second-order. By utilizing the state information of multi-order neighbors, a novel consensus protocol is proposed. Compared to the existing controllers that use only first-order neighbors’ information, or both first-order and second-order neighbors’ information, the controller we proposed can accelerate convergence of the system by simply increasing the value of the parameter $g$. Sufficient and necessary conditions are obtained for consensus of the systems with the proposed protocols. Meanwhile, a dynamic event-triggered mechanism is designed in order to reduce the communication frequency among agents to save communication resources. As the parameter $g$ increases, the frequency of events being triggered decreases. The validity of the theoretical results is verified by simulations.