Intelligent Adaptive Control Method for Ducted Fan Unmanned Aircraft

Ducted fan vertical take-off and landing unmanned aerial vehicle has a compact layout and can adapt to various flight missions under complex flight conditions. However, the aerodynamic characteristics of ducted fan vehicle are complex and difficult to model accurately, so applying conventional flight control methods to ducted fan vehicle cannot achieve the desired control performance and robustness. In this paper, an intelligent adaptive control design method using dynamic inversion combined with on-line neural network is developed for a ducted fan vehicle. This method requires only the linear dynamic model of the vehicle at hovering state to achieve hovering and low-speed flight control of ducted fan vehicle by utilizing the adaptive error compensation capability of the neural network. The proposed method greatly reduces the requirement for the accurate dynamic model of the UAV, and finally the effectiveness of the method is verified by flight simulation.