Saturation-tolerant prescribed performance neural formation control for air-floating robots under false data injection attacks

This paper investigates the formation control problem for air-floating robot (AFR) systems, accounting for input saturation and false data injection (FDI) attacks. A confidence-factor-augmented distributed observer is designed to reconstruct leader motion states under partial observability constraints, actively mitigating neighbor-induced uncertainties in AFR swarms. Furthermore, by integrating neural networks with an extended state observer, the proposed distributed controller achieves disturbance estimation and compensation for desired formation configuration. To address static constraint limitations, a saturation-tolerant prescribed performance controller leverages an auxiliary system that adaptively governs dynamic tracking boundaries, effectively resolving intrinsic brittleness as well as actuator failures and saturation problems.Theoretical analysis guarantees system stability, with experimental results demonstrating the method’s effectiveness.