Neuro-adaptive Containment of Uncertain Complex Cyber Physical Networks with Directed Topology

Abstract

This paper studies the containment problem for complex cyber-physical networks (CCPNs) subject to parameter uncertainties and external disturbances. By using the neural network (NN) approximation theory, a continuous neuro-adaptive containment controller is designed, where the NN adaptive law is used to adjust the NN weights and the other adaptive laws are used to adjust the network coupling strengths. And we prove that the containment error is uniformly ultimately bounded (UUB) if the graph among followers is detailed balanced and for each follower, there exists at least one leader has a directed path to it. As the containment criteria depend only on local information, the achieved containment is fully distributed. A favourable property of the containment controller is chattering free since it is continuous. Finally, the theoretical result is validated by numerical simulation.