基于模型拼接技术的四旋翼无人机全飞行包线建模

Quadrotor vehicles have found wide application in both military and civil fields. However, the onboard sensors of quadrotor vehicles are poor in reliability and accuracy, and the dynamic models of quadrotors vary considerably under different flight conditions. This study systematically examines the flight path reconstruction, frequency domain system identification, and full flight envelope modeling of quadrotor vehicles. To overcome the shortcomings of significant constant bias errors and unknown statistical characteristics of measurement noise of the airborne low-cost sensors, we use the flight path reconstruction technology based on Extended Kalman Filter(EKF) to reconstruct the airspeed of UAVs in flight tests. The frequency domain method is adopted to identify the body dynamics model under different flight conditions aiming at the different characteristics in dynamic models of quadrotor UAVs with different flight speeds. In view of the high-precision control requirements of the four rotor UAVs in large-span areas, a systematic full flight envelope modeling scheme based on the model stitching technology is proposed, which is verified to be accurate and reliable.