Neural network-based robust fully distributed event-triggered formation control for uncertain nonlinear multiagent systems

This research focuses on the distributed formation issue of a class of uncertain nonlinear multiagent systems. Specifically, the objective system suffers from unknown nonlinear dynamics, external disturbances and parameter uncertainties. With the aid of adaptive technique, an ideal fully distributed event-triggered formation reference signal without requiring any knowledge of global information is first designed. Then an RBF neural network (NN) is employed to approximate the nonlinear dynamics. On the basis of the work above, an adaptive NN-based robust event-triggered controller is further delicately constructed, which can guarantee that the formation goal is achieved. A distinct feature of this robust controller, compared to related works, is its ability to completely eliminate the adverse effects caused by disturbances/uncertainties without inducing chattering. Moreover, the proposed scheme can not only achieve discrete information interactions between neighboring agents like most existing articles but also enable discrete updates of the controller. Finally, the theoretical results are demonstrated via a numerical example.