Wenz, Andreas Wolfgang (Norwegian University of Science and Technology), Johansen, Tor Arne (Norweigian Univ. Of Sci. & Tech.)

Estimation of Wind Velocities and Aerodynamic Coefficients for UAVs Using Standard Autopilot Sensors and a Moving Horizon Estimator

Scheduled for presentation during the "Intelligent and Autonomous UASs" (ThC5), Thursday, June 15, 2017, 17:00−17:20, San Marco Island

2017 International Conference on Unmanned Aircraft Systems, June 13-16, 2017, Miami Marriott Biscayne Bay, Miami, FL,

This information is tentative and subject to change. Compiled on April 24, 2024

Abstract

While operating any aircraft it is vital to know its current flight state. Some of the most important variables to assess the flight state are the airspeed, the angle of attack and the sideslip angle. Larger aircraft are equipped with sensors specifically designed to measure these variables. How- ever on small unmanned aerial vehicles (UAVs) much stricter restrictions on size, weight and cost prohibit the use of such sensors. Therefore we propose a method to estimate the airflow variables utilizing only sensors that are part of a standard UAV autopilot. This includes an inertial measurement unit (IMU), a global navigation satellite system (GNSS) receiver and a pitot-static tube. These measurement together with kinematic and aerodynamic models will be fused within an estimator to estimate steady and turbulent wind velocities as well as aerodynamic coefficients. With these estimates it is possible to calculate the angle of attack, the sideslip angle and the airspeed. A main challenge is to distinguish between changes in the aerodynamic coefficients and changes in wind velocity, since pitot-static tube measurements of the relative airspeed are only available in one direction at a time and hence the system is not always observable. Therefore attitude changes have to be undertaken to achieve persistence of excitation. In this paper a Moving Horizon Estimator (MHE) is used for estimation. Simulation results show overall good estimation results and significant improvements compared to a previous Extended Kalman Filter approach. Root mean square errors (RMSE) are 0.25 degrees for the angle of attack, 0.08m/s for the airspeed and 1.06 degrees for the side slip estimates.