Neural network-based reinforcement learning for prescribed-time formation tracking of hybrid networked marine systems under attacks.

This paper focuses on solving the time-varying formation tracking (TVFT) problem for surface-underwater hybrid networked marine systems (HNMSs) under denial of service (DoS) attacks and physical attacks. The system comprises multiple autonomous surface vehicles (ASVs) and autonomous underwater vehicles (AUVs). To address this challenge, a prescribed-time hierarchical control (PTHC) framework is proposed. For each follower, a prescribed-time distributed estimator is developed to estimate the states of leaders using relative information from neighboring agents, even in the presence of DoS attacks. To mitigate the effects of physical attacks, external disturbances, and unmodeled hydrodynamics, a reinforcement learning algorithm is incorporated into a neural network-based local controller, enabling time-varying formation tracking within a prescribed time. Using Lyapunov stability theory, sufficient conditions are established to guarantee the prescribed-time stability of the closed-loop system. Finally, simulations are conducted to validate the effectiveness and practicality of the proposed approach.