Automatized deck landing for a double rotors helicopter UAV

The prime objective of this research would be to conduct autonomous precision landing of an Unmanned Autonomous Vehicle (UAV) on mobile platforms including maritime platform and military vessel helipad. The developed system will be used on LX300, a real and considerable dimensions twin-rotor unmanned helicopter, manufactured by Laflamme Aero and with the goal of maritime surveillance. This UAV will navigate in area where there are no navigation beacons and will only be able to use satellite systems. However, maritime decks are small and can’t handle traditional beacon system because of their structure and are subject to the strong swell which cause big attitude changes. Hence, this matter necessitates better alternatives to get a better accuracy than a GNSS system, and the synchronization of the drone and the platform to get a soft, accurate and safe landing for both the drone and the platform.
To materialize this goal, two parallel systems will be developed to compute the relative distance and the relative attitude between the drone and the platform:
- The first consists of a robust system of tracking the position of the drone in relation to the platform, calculating the range and synchronizing attitude by communicating between two little sized modules, one on the platform and the other on the drone. A GPS/Real Time Kinematic (RTK) system will be used for relative position and a relative IMU/AHRS system to synchronise attitudes.
- The second method is to use a lightweight vision system along with four rangefinders. The First phase would be detection of the landing zone from a certain distance. Then, the trajectory will be built, and the control action will be computed. In succeeding stage, the camera will keep the rotorcraft oriented with ship landing deck and simultaneously the rangefinders will model the attitude of the landing zone. Lastly, the most optimized and safe time will be calculated, and the final alighting task will be conducted.
Finally, to approach a more effective system these sensors data will be fused, using filters like extended Kalman filter and unscented Kalman filters.