Airflow Angles Estimation-Based Finite-Time Adaptive Neural Control for Aircraft at High-Angle-of-Attack Maneuvers

This article investigates the high-angle-of-attack (high-AOA) maneuver control problem for aircraft through finite-time techniques and neural learning. Considering the performance degradation of the flush air data sensing system at high-AOA maneuvers, a Kalman filtering approach based on force equations is implemented to estimate airflow angles online using signals from the inertial navigation system, even when aerodynamic coefficients are unknown. With the filtered signals, a finite-time adaptive neural controller is developed to generate the desired control moment and achieve rapid tracking of high-AOA commands, where both neural network (NN) and disturbance observer (DOB) are integrated to estimate composite disturbances. To enhance learning performance, an adaptive evaluation signal is constructed using online recorded data to update NN and DOB parameters. The deflections of thrust vector nozzles and aerodynamic control surfaces are finally obtained by solving an optimal control allocation problem. The practical finite-time uniformly ultimately bounded stability is proved through Lyapunov analysis. Herbst Maneuver simulations demonstrate that the proposed design achieves superior performance in both tracking accuracy and learning capabilities.