Research on Flight Control of TRUAVs Based on Adaptive Dynamic Programming

This paper addresses the challenge of ensuring stable flight for tilt-rotor unmanned aerial vehicles (TRUAVs) in the presence of parameter perturbations and environmental disturbances. Based on the given mathematical model of TRUAVs, a novel flight control framework is designed to achieve integrated position and attitude control for the hovering mode in a hybrid control scheme. Subsequently, to accommodate the structure and environmental changes, a data-driven ADHDP controller that is capable of adjusting the control strategy online is proposed. In the ADHDP controller, BP neural network is employed to build the actor and critic networks, and algorithm flow is also given. Simulation experiments show that the controller's robustness against structure and environmental disturbances has been proved comparing with the traditional PID controller.