Remote Guidance Method of Unmanned Aerial Vehicle Based on Multi-sensors

Aiming at the problem that single sensor cannot balance the distance of guidance and accuracy, the paper proposes a remote guidance method for UAV (Unmanned Aerial Vehicle) based on multi-sensors. This paper first introduces the remote guidance and autonomous landing task. Then the hardware and control method are designed. The landing process is divided into two stages: the long-range and close-range. In order to take advantage of the characteristics of different sensors, GPS/IMU (Inertial Measurement Unit) hybrid guidance is used for long-range stage and visual guidance is used for close-range stage. The positioning and guidance method are also introduced in detail. In close- range stage, according to the requirements of accurate landing on mobile UGV (Unmanned Ground Vehicle), the vision guidance method based on Apriltag is proposed and the calculation of UAV attitude and relative distance is completed. EKF (Extended Kalman Filter) is used as the multi-sensors fusion method. Finally, the method is applied to a quadrotor and Mecanum UGV and an outdoor experiment is realized and analyzed. The experimental results show that the accuracy and real-time ability of the proposed method meet the requirements of UAV’s landing on a mobile UGV from a long distance.