Modeling and Neuroadaptive Output Feedback Attitude Control for Receiver UAV Under State Constraints

This paper addresses the output feedback attitude tracking control problem for a receiver unmanned aerial vehicle (UAV) with state constraints during the refueling phase of autonomous aerial refueling (AAR). First, dynamic models of the receiver UAV, affected by fuel injection and external airflow disturbances, are established, considering the time-varying mass, centroid, inertia, and incremental moments induced by the abovementioned factors. Next, leveraging a self-improving double recurrent fuzzy neural network (DRFNN), an adaptive fuzzy neural network observer (AFNNO) is developed to estimate the angular velocity. The DRFNN is employed to approximate the system's unknown dynamics, and its self-improving mechanism optimizes the number of rules based on rule similarity and reasonability, thereby enhancing approximation accuracy. Following this, a novel asymmetric barrier Lyapunov function (BLF), developed using a hyperbolic tangent function method to avoid piecewise functions, facilitates a backstepping controller to achieve constrained output. Finally, numerical simulations are presented to validate the effectiveness of the proposed control scheme.