Model-Free Integrated Navigation of Small Fixed-Wing UAVs Full State Estimation in Wind Disturbance

This paper presents a model-free distributed multi-sensor extended Kalman filter (DMSEKF) full state estimation algorithm to provide long-term convergent flight parameters for small fixed-wing unmanned aerial vehicles (UAVs). The full state has the attitude, velocity, position, airspeed, and 2D horizontal wind speed. The airspeed and wind speed are estimated in wind disturbance to provide more robust perception information. The model-free estimator has a low-cost standard sensor suite, including an IMU, a magnetometer, a barometer, a GPS module, and an airspeed tube, rather than the aerodynamic model of the UAVs to increase the multi-sensor fusion algorithm versatility in various UAVs. Then, the full state integrated navigation model is established based on the onboard sensor suite fused by the distributed tightly-coupled EKF. In addition, a consistent multiple sensors data processing method is designed to synchronize the time node of all onboard sensors. Finally, the proposed algorithm is verified through the experimental flight sensor data. The results demonstrate that the proposed algorithm can provide a reliable full state vector and achieve an effective solution performance during the UAVs flight application.

Automated summary

This paper presents a model-free distributed multi-sensor extended Kalman filter algorithm for small fixed-wing unmanned aerial vehicles to provide long-term convergent flight parameters. The algorithm estimates airspeed and wind speed in wind disturbance to improve perception information robustness. The algorithm utilizes a low-cost standard sensor suite to increase versatility. Experimental results demonstrate the reliability and effectiveness of the algorithm.