Data-Driven Distributed Adaptive Consensus Tracking of Nonlinear Multiagent Systems: A Controller-Based Dynamic Linearization Method
Xian Yu, Zhongsheng Hou, Tianshi Chen
Abstract
For the consensus tracking of multiagent systems (MASs), most of existing distributed control methods need to design the controller structure with availability of physical model or structural information of each agent, which is sometimes impractical due to the difficulty of modeling each agent or obtaining its structural information if the physical model of each agent is complicated and the underlined MAS is heterogeneous. To handle this issue, in this article we first integrate the controller-based dynamic linearization method into distributed control using only the local measurement information exchanging among neighbors in a directional graph and the input of each agent. Then we propose a data-driven distributed adaptive control (DAC) method for nonlinear nonaffine heterogeneous discrete-time leader–follower MASs. We show that the proposed method has ultimately bounded tracking error in both of the cases with fixed and switching communication topologies. The numerical simulation results show that in contrast with two DAC methods, the proposed one can give smaller tracking errors.