Neural adaptive input-output constrained control for the longitudinal dynamics of UAV with an input delay

Achieving stable and precise trajectory tracking for the longitudinal dynamics of unmanned aerial vehicles (UAVs) with input delays and external uncertain disturbances is particularly challenging due to the high-frequency state oscillations and the transient control input jump. This paper proposes a novel delayed feedback neural adaptive input-output constrained control (DF-NAIOCC) strategy to significantly reduce the state oscillations and control jump, enhancing the overall operational smoothness of the system. By extending the delayed feedback neural adaptive control (DF-NAC) strategy based on radial basis function neural networks (RBFNNs) with the integration of a barrier Lyapunov function (BLF), the delayed feedback neural adaptive output-constrained control (DF-NAOCC) strategy is formulated. Further, a neurodynamic model is integrated with DF-NAOCC to develop the proposed DF-NAIOCC strategy. The predictive states serve as approximations of the future states in the control strategies. Comprehensive simulations of the trajectory tracking control of a UAV’s longitudinal dynamic system with an input delay and external disturbances validate the theoretical analysis, demonstrating substantial enhancements in tracking accuracy, estimation accuracy, and state smoothness achieved by the DF-NAIOCC compared to the baseline DF-NAC and DF-NAOCC.