Prescribed Performance-Based Consensus of Nonlinear Multiagent Systems With Unknown Control Directions and Switching Networks.

In this paper, the prescribed performance-based consensus problem for nonlinear multiagent systems with unknown control directions and switching networks is investigated. Different from the existing works, the unknown control directions and prescribed performance of consensus errors are considered simultaneously under the switching networks. In this case, the graph Laplacian matrix may be singular and the controllers are interacted. To handle this problem which cannot be dealt with directly using the existing methods, a distributed adaptive observer is introduced so that the controllers are decoupled and the difficulty caused by the unknown control directions can be overcome by adopting the existing Nussbaum functions and the unknown control directions of agents are allowed to be nonidentical. Meanwhile, by utilizing the designed controller, the consensus errors can be guaranteed with the prescribed performance. Finally, a numerical example verifies the validity of the proposed method. [ABSTRACT FROM AUTHOR]