Hierarchical hybrid event-triggered neural control for air-bearing robot formation with prescribed performance and collision avoidance

This paper proposes a hierarchical adaptive control architecture with flexible prescribed performance for consensus tracking in air bearing robots (ABR) formations under constrained communications and external disturbances. A hierarchical control framework is proposed, which interconnects leaders and followers through a virtual system, preventing collisions among air-bearing robots through upper-layer prescribed performance parameter design while blocking mutual propagation of disturbance or fault signals. To address unmeasurable velocity and unknown disturbances, a novel neural network based extended state observer is synthesized, which using one algebraic iteration as the iterative learning algorithm to reduce computational complexity. Furthermore, a saturation threshold hybrid triggering strategy with transient performance guarantees is proposed, effectively reducing communication overhead by 48% while preventing actuator saturation-induced fragility in multi-constraint scenarios. Theoretical analysis guarantees system stability, and experimental results demonstrate the method's effectiveness.