Abstract
Three recently published attitude estimation algorithms are compared, and another one is proposed, aiming the performance evaluation when used in an attitude and heading reference system (AHRS), using low cost MEMS sensors, for fixed wing Unmanned Aerial Vehicles. The comparison is based on simulation results associated with typical aerial maneuvers of fixed wing UAVs, characterized by low frequency acceleration signals that are hard to be distinguished from intrinsic sensors biases, as it is usual during coordinated turns. The sensors models are also parameterized from information obtained in the respective datasheets, or obtained from experimental procedures, in order to better represent the imperfections present in practice. Three algorithms are based on the EKF (Extended Kalman Filter) and one is based on nonlinear complementary filtering. The results have revealed that some form of compensation for the effect of low frequency accelerations seems to be crucial to achieve good results.