Consensus tracking control for a class of general linear hybrid multi-agent systems: A model-free approach

This paper studies the consensus tracking problem for a class of general linear hybrid multi-agent systems via a model-free approach. The hybrid multi-agent system is composed of N followers with discrete-time or continuous-time linear dynamics and one leader with discrete-time linear dynamics. The proposed distributed consensus tracking controller consists of a compensation term and an optimal control term. Specifically, the compensation term is introduced to address the effect of agents' inherent dynamics on consensus tracking, and the optimal control term is designed to achieve the performance goal. By designing distributed observers for the following agents, we first propose a model-based distributed consensus tracking protocol. A model-free distributed consensus tracking protocol is further constructed via designing adaptive observers and identifying agents' dynamics. The salient features of the model-free consensus tracking protocol are that the system matrix of the leader is only available for some followers and any follower does not need to know the eigenvalues of the graph matrix. Moreover, in order to reduce the computational burden, a system identification process is implemented based on the parallel learning method. Finally, the presented theoretical results are verified by simulation.