Sensor Fusion for Quadrotor Autonomous Navigation

摘要

Autonomous navigation requires the availability of all vehicle states. Since there is not a single sensor capable of measuring translational and rotational positions and velocities, sensor fusion becomes necessary. This paper proposes a translational velocity observer and a position scale factor estimator based on inertial and visual measurements. The observer/estimator design follows the Immersion and Invariance method. Numerical simulations and real-time experiments are presented to validate the proposed observer/estimator.