Cooperative Trajectory Tracking of Heterogeneous Networks by Distributed Learning

This article presents a distributed learning approach to achieve the high-precision cooperative trajectory tracking tasks for heterogeneous networks at all time in the presence of changing topologies. An updating law of distributed learning is proposed by leveraging the cooperative tracking error trajectories of agents at all time samples, thanks to which a monotonic convergence result is established for heterogeneous networks of linear agents without iteration-varying uncertainties. When considering heterogeneous networks of nonlinear agents with iteration-varying uncertainties, robust convergence results are further explored such that not only can robust cooperative trajectory tracking be realized, but also a robust stability property can be acquired for distributed learning processes. To carry out robust convergence analysis of distributed learning, an extended contraction mapping and a heterogeneous-to-homogeneous transformation approaches are developed, which can particularly address full heterogeneities and unknown nonlinearities of agents.