Prediction of the Navigation Angles Using Random Forest Algorithm and Real Flight Data of UAVs

Abstract

Unmanned aerial vehicles (UAVs) are becoming more common in aviation. Predicting and estimating of UAV attitude are important aspects of the communication strategy. Autonomous UAVs' attitude estimate is a crucial component for controlling in flight mission. The attitude can take many forms such as Euler angles and Quaternion. This paper explains attitude estimation of UAV using Random Forest (RF) algorithm based on real flight data of Accelerations, Angular velocity, Magnetic Field, and GPS measurements for prediction current attitude angles (roll, pitch and yaw). The developed algorithm uses features that are derives from sensors measurements. The developed RF model results showed accurate prediction of UAVs attitude angles with small error values.