Consensus analysis of multiagent systems with second-order nonlinear dynamics and general directed topology: An event-triggered scheme

Event-triggered sampling control is motivated by the application of embedded microprocessors equipped in the agents with limited computation and storage resources. This paper studies the global consensus in second-order multi-agent systems with the inherent nonlinear dynamics on general directed networks using decentralized event-triggered strategy. For each agent, only utilizing local and current sampling data, the update of controllers is event-based and only triggered at their own event times. A high-performance sampling event that only needs neighbors' states at their own discrete time instants is presented. Furthermore, we introduce two kinds of general algebraic connectivity for strongly connected networks and strongly connected components of directed networks containing a spanning tree to describe the system's ability to reach consensus. A detailed theoretical analysis on consensus is performed and two criteria are derived by the virtues of algebraic graph theory, matrix theory, and Lyapunov control approach. It is shown that the continuous communication between neighboring agents can be avoided and the Zeno-behavior of triggered time sequence is excluded during the system's whole working process. In addition, numerical simulation is given to illustrate the effectiveness of the theoretical results.