Design of the UAV Trajectory Tracking System Based on the Adaptive Neural Network - ADRC Method

In order to solve the problems of internal uncertainty and external disturbance of an unmanned aerial vehicle (UAV), this paper proposes the active disturbance rejection controller (ADRC) based on the adaptive radial basis function (RBF) neural network. Firstly, the dynamic model of a quadrotor with disturbance is analyzed and the controller is designed based on the ADRC method. Secondly, the RBF network is used to estimate the unknown parameter $b$ of the system and the Lyapunov function is constructed to prove the stability of the closed-loop system. Finally, simulation results of a spiral ascent and a climbing maneuver flight illustrate that the controller proposed in this paper has high tracking accuracy and strong robustness under different flight scenarios; the average tracking error of the outdoor flight experiment is about 0.22 m, which further verifies the effectiveness of the proposed controller.