Distributed Control of Nonlinear Multiagent Systems With Unknown and Nonidentical Control Directions via Event-Triggered Communication

• Published in: IEEE transactions on cybernetics Vol. 50; no. 5; pp. 1820 - 1832
• Main Authors: Wang, Yan-Wu, Lei, Yan, Bian, Tao, Guan, Zhi-Hong
• Format: Journal Article
• Language: English
• Published: United States IEEE 01.05.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Communication; Control systems; Directed interaction topology; event-triggered communication; Internal model principle; jointly connected; Laplace equations; Monitoring; Multi-agent systems; Multiagent systems; Nonlinear control; nonlinear multiagent system (MAS); Nonlinear systems; Parameter uncertainty; Switches; Switching theory; Symmetric matrices; Topology; unknown nonidentical control direction
• ISSN: 2168-2267; 2168-2275; 2168-2275
• DOI: 10.1109/TCYB.2019.2908874

In this paper, the leader-following output consensus problem for a class of uncertain nonlinear multiagent systems with unknown control directions is investigated. Each agent system has nonidentical dynamics and is subject to external disturbances and uncertain parameters. The agents are connected through a directed and jointly connected switching network. A novel two-layer distributed hierarchical control scheme is proposed. In the upper layer, to save the communication resources and to handle the switching networks, an event-triggered communication scheme is proposed, and a Zeno-free event-triggered mechanism is designed for each agent to generate the asynchronous triggering time instants. Furthermore, to avoid the continuous monitoring of the system states, a Zeno-free self-triggering algorithm is proposed. In the lower layer, to handle the unknown control directions problem and to achieve the output tracking of the local references generated in the upper layer, the Nussbaum-type function-based technique is combined with internal model principle. With the proposed two-layer distributed hierarchical controller, the leader-following output consensus is achieved. The obtained result is further extended to the formation control problem. Finally, three numerical examples are provided to demonstrate the effectiveness of the proposed theoretical results.