Consensus of multi-agent nonlinear systems by directed interactions: An internal model approach

This paper considers a global robust output consensus problem for multi-agent nonlinear systems by general directed output interactions. For heterogeneous and minimum-phase nonlinear agents transformable into an output-feedback form, it is shown that the problem is solvable by an internal model approach under certain mild conditions. A dynamic output-feedback protocol is developed by converting the output consensus into a distributed stabilization of an augmented dynamic network.