Consensus of heterogeneous multi-agent nonlinear systems with unknown leaders

This paper considers a global robust output consensus (GROC) problem for heterogeneous multi-agent nonlinear systems with unknown leaders. For a class of normal-form nonlinear agents under general directed communication topologies, we develop a new nonlinear internal model approach for the problem. Specifically, by means of an effective problem conversion, we can establish an augmented dynamic network with integral input-to-state stable (iISS) inverse dynamics. Moreover, we are capable of constructing an output feedback consensus protocol through a direct auxiliary distributed stabilization design.