Ship deck landing of a quadrotor using the invariant ellipsoid technique

Vertical takeoff and landing (VTOL) unmanned aerial vehicle (UAV) operations on ships enhance the mission capabilities for military and civilian users. However, landing operations on a ship requires precise control to avoid collision with the ship superstructures. Current approach involves modelling and prediction of the ship deck motion. However, model mismatch often results in a non-zero bounded error response. The approach in this paper eliminates modelling of the ship deck motion and focuses on an analysis method based on the invariant ellipsoid technique to obtain an optimal gain such that the tracking error is restricted to within an optimal bound. Simulations are performed using a realistic ship deck model with the quadrotor tracking a point at a prescribed height above the ship deck before landing. The performance of the proposed controller on a quadrotor was evaluated and is shown to achieve the control objective.