A REAL-TIME ATTITUDE ESTIMATION SCHEME FOR HEXAROTOR MIC RO AERIAL VEHICLE

Abstract. Position and attitude estimation is vital for vertical take-off and landing unmanned aerial vehicle ( VTOL-UAV) that can be executed by integrating the output of gyros, accelerometers and magnetometers with respectively gravity and local magnetic field vectors. For this purpose, a Kalman filter with different variant as extended, unscented and complementary has been largely used in the literature. Another technique, vision data to navigate an unknown, indoor, GPS-denied environment with optical flow or visual servo based on image or position was used. Without external sensing, an estimation system that relies only on integrating inertial data will have rapidly drifting position estimates. VTOL-UAVs are stringently weight constrained, leaving little margin for additional sensors beyond the mission payload. The aim of this paper is to present a realtime sensor fusion scheme based on nonlinear filtering, for hexarotor UAV to localization problem. The absolute attitude estimation use the extended Kalman filter based on quaternion to avoid singularities. Results obtained in realtime system to the hexarotor UAV shows the better attitude performance.