Secure Formation Tracking of Networked MSVs Under Actuator Constraints and DoS Attacks: Data-Driven Observer-based Control

This paper studies the secure formation tracking (SFT) problem of networked marine surface vehicles (NMSVs) in the presence of denial of service (DoS) attacks, model uncertainties, external disturbances and actuator constraints. A cyberphysical framework is developed for designing the data-driven observer-based control (DDOBC) algorithm, which consists of two layers. In the cyber layer, a distributed resilient trajectory estimator is proposed to accurately estimate the information of virtual leader under DoS attacks over digraphs. In the physical layer, a neural network-based (NN-based) data-driven observer is firstly utilized to resolve the model uncertainties. Then, a observer-based local controller is designed to achieve the SFT problem with input saturations and actuator faults. The sufficient conditions for system convergence and stability are derived based on the Lyapunov stability theory. Finally, simulation experiments are conducted to verify the effectiveness of the theoretical results.