Fully Distributed Optimal Tracking Control for UAV-UGV Formation Against Byzantine Attacks

This paper explores the issue of fully distributed optimal tracking control of unmanned aerial vehicle (UAV) and unmanned ground vehicle (UGV) formation under Byzantine attacks. Based on a novel constructed second-order nonlinear UAV-UGV swarm in space, the leader's dynamics are presumed to be linear in relation to its state variables, with the coefficient matrices remaining unknown to all followers. Meanwhile, considering that the UAV-UGV swarm is susceptible to the propagation of incorrect neighbors' information and false input signals (called Byzantine attacks), which is effectively reinterpreted as the management of unknown variables within the control inputs, an optimal control method under the reinforcement learning (RL) algorithm is investigated to against the aggressive Byzantine attacks. Furthermore, the neural network (NN) is employed to approximate the Byzantine attacks and unknown terms arising from the leader's dynamics and Hamilton-Jacobi-Bellman (HJB) equation, and the actor and critic adaptive laws are designed in actor-critic-identifier architecture. Subsequently, an optimal tracking control scheme is proposed to ensure fully distributed collaborative tracking of the UAV-UGV formation under Byzantine attacks. Finally, simulation and experimental results are carried out to demonstrate the efficacy of the proposed approach.