A Performance Region‐based Approach to the Leader‐following Consensus of Nonlinear Multiagent Systems
Jingyao Wang,Zhisheng Duan
DOI: https://doi.org/10.1002/rnc.5370
IF: 3.8973
2021-01-01
International Journal of Robust and Nonlinear Control
Abstract:The communication network is usually susceptible to effective communication channel capacity variation and transmission imperfections, which renders the connection among networked agents unstable. Motivated by this fact, we assume that the communication graphs are Markovian randomly switching, because the Markovian chain process is commonly employed to describe the network subject to Rayleigh fading channel. Under such a communication connection assumption, this article investigates the globally Script capital H-infinity leader-following consensus control problem for grouped agents with nonlinear dynamics. Different nonlinear dynamics, including Lur'e and Lorenz-type chaotic systems, are considered. By using the tools from the robust control theory, the conditions guaranteeing the globally Script capital H-infinity leader-following consensus are derived. Under such conditions, consensus control protocols can be designed solely based on the relative state information transmitted over the communication network, by solving some linear matrix inequalities. Then, we characterize the robustness of the controlled systems against the variation of the communication graphs with the globally Script capital H-infinity leader-following consensus performance region. It is theoretically shown that the networked nonlinear systems under the proposed controller yield unbounded and connected robust performance regions. Finally, the theoretical results are verified by conducting numerical simulations on networks consisting of Chua's oscillators and Lu systems.