Non-linear model predictive control for guidance of a fixed-wing UAV in precision deep stall landing

In order to achieve a high degree of operational flexibility, it is often required to recover small UAVs without infrastructure such as a runway. To help overcome this difficulty, deep stall landing can be used as a landing method on small space. The UAV is in a deep stall when the angle of attack is beyond the stall angle, and the condition causes the UAV to loose height rapidly. In the post-stall phase the angle of attack is controlled and the UAV can land with relatively low speed. In this article we focus on the highly non-linear longitudinal dynamics, and employ a non-linear model predictive control (MPC) to find a control sequence to guide a model of a fixed-wing UAV into a deep stall to land at a given location and given path angle with minimum speed.