Development of autonomous quadrotor system for vertical replenishment

Cargo transportation between moving platforms is one of the most meaningful potential applications of UAVs. However, to fully autonomously transport the cargos, the UAV is required to have high position precision and high intelligence to tackle all the emergencies, on which the technology is not mature yet. In this article, we present a quadrotor system targeting at fully autonomous cargo transportation missions. A large payload, long endurance quadorotor is utilized as the UAV platform. A novel navigation framework is proposed for the UAV to navigate between moving platforms. A robust vision algorithm is presented for localizing visual target under changing illumination conditions with auto-adjustment of exposure and HSV values. Sophisticated mission planning algorithm is developed with highly modular software structure for managing the complex mission requirements and tackling different emergencies. The developed quadrotor system has taken part in the 3rd AVIC Cup — International UAV Innovation Grand Prix 2015 and our team won the championship.