Funnel-Based Adaptive Optimal Resilient Control for Air–Ground Cooperative Systems With Composite Disturbances

This article addresses the adaptive neural network-based optimal resilient control problem for a class of air–ground cooperative systems subject to the false data injection (FDI) attacks and the composite disturbances. First, a novel funnel boundary is designed to constrain formation errors with prescribed transient and steady-state performance, ensuring collision avoidance. Second, a composite disturbance observer (CDO) is proposed to accurately estimate and compensate for both external and neighbor-induced coupled disturbances. Furthermore, an optimal resilient control strategy is formulated within a backstepping framework with an actor–critic neural network (NN) to mitigate FDI attacks, ensuring stable formation tracking and performance optimization. The theoretical analysis proves the boundedness of all closed-loop signals and the convergence of formation errors within the funnel. The numerical simulations validate the approach’s effectiveness and superiority.