Node‐to‐node consensus of multi‐agent networks with event‐triggered control and packet losses

Abstract

This paper studies the node‐to‐node consensus problem for multi‐agent networks possessing a leaders' layer and a followers' layer via the pinning control. In order to realize the consensus and reduce the update frequency of the controller, a suitable event‐triggered mechanism is introduced into the control strategy. Furthermore, the phenomenon of packet loss is considered in the designed controller. Based on the M‐matrix theory and Lyapunov stability theory, this paper presents the sufficient conditions for the node‐to‐node consensus of networks. Meanwhile, it is proved that the Zeno behaviour is excluded. Finally, two numerical simulations are provided to demonstrate the effectiveness of the theoretical results.

Synthesis andL2‐gain analysis for switched systems under event‐triggered switching

“…With the development of communication network, event-triggered control becomes the focus of many researches in recent years [25][26][27]. Qi et al [28] investigated the event-triggered control for network-based switched linear systems with transmission delay.…”

Node‐to‐node bipartite consensus of multi‐agent systems with disturbances

“…The leaders' layer contains multiple interactive virtual leaders, and a small fraction of the followers in the followers' layer are controlled by their corresponding leaders, then all of the followers will track their corresponding leaders. To address this phenomenon, the concept of node‐to‐node consensus has been put forward and investigated [14–17]. Given that only a part of followers were pinned by their leaders with time‐varying pinning links, some criteria for node‐to‐node consensus were provided in [14].…”

“…By utilising the sampled‐data control, node‐to‐node consensus was realised in directed multi‐agent networks with non‐linear dynamics [16]. In addition, our recent work focused on node‐to‐node consensus under an event‐triggered strategy [17].…”

“…It is worth pointing out that neighbouring agents are supposed to have only cooperative interactions in the aforementioned works [14–17]. However, the relationship between agents may be competitive besides cooperative.…”

“…This paper further focuses on the consensus behaviour for multi‐agent systems with antagonistic and cooperative relationships. Motivated by the existing results [14–17], we suppose that the multi‐agent system consists of two layers, i.e. a leaders' layer and a followers' layer.…”