Approach maneuvers for autonomous landing using visual servo control

A nonlinear image-based visual servo control algorithm for autonomous landing of a fixed wing aircraft is described. The primary sensor system is a vision sensor yielding a sequence of images from which 2D linear and point features of the runway are extracted. The first two phases of a landing maneuver, alignment to the runway and glide (descent to the runway) are treated in the work presented here. The final landing maneuvers, flare to touchdown, and taxiing require additional sensor modalities and are not treated. The proposed control scheme deals with unknown wind conditions and incorporates the full nonlinear dynamics of the airplane. Simulation results based on realistic environmental conditions and measurement noise are presented that validate the control design approach.