UAV Attitude and Heading Algorithm Based on Extended Kalman Filter

Aiming at the problem of low attitude resolution accuracy of small unmanned aerial vehicle（UAV） in large maneuver or continuous flight, an attitude algorithm based on extended Kalman filter（EKF） was proposed. In this algorithm, the acceleration component of the aircraft system and gyro drift were taken as the state variables to be estimated, and a nonlinear filtering model was established. Then, the measured values of sensors were preprocessed and data fusion was completed to obtain the accurate output of attitude angle data. At the same time, the noise covariance was constantly adjusted according to the external motion acceleration to achieve the adaptive correction of EKF, and the influence of magnetic interference in UAV attitude calculation was suppressed. Through three-axis turntable experimental test, the static error of attitude angle is not more than 0.5°, the dynamic error is not more than 2°, and the transient magnetic interference error is not more than 5°. It is shown that this algorithm can significantly improve the attitude calculation accuracy of small UAV.