Application of AEKF Parameter Identification to UAV Autopilot

摘要

The autopilot of a UAV(Unmanned Aerial Vehicle) is designed, and the online aerodynamic parameter identification based on AEKF is applied to the autopilot in this paper. The autopilot is designed according to classical control theories at first. Then, the AEKF (Augmented Extended Kalman Filter) with the strong tracking algorithm is applied to the autopilot, to identify the important aerodynamic parameters in closed loop. The autopilot calculates control gains using the aerodynamic parameters identified by AEKF, the flight velocity, and the flight altitude, to form an adaptive autopilot. The six degree of freedom simulation indicates that during flight, the AEKF, which works in closed loop, is capable of identifying the important aerodynamic parameters accurately using the measured fin-deflexions and the IMU data. This adaptive autopilot based on AEKF is able to overcome large prediction error, or even abrupt change of the aerodynamic parameters. This autopilot is in possession of strong robustness.