Game theoretic collision avoidance algorithms for small UAV

Abstract

Unmanned aerial vehicle refers to aircraft that are remotely controlled from the ground or autonomously fly according to a pre-programmed route without a pilot aboard. In particular, demand for small unmanned aerial vehicle is continuously increasing in the public and private sectors. Furthermore, small unmanned aerial vehicles are operated in low-altitude and uncontrolled airspace, which are not supported by air traffic control. Accordingly, practical operation of unmanned aerial vehicles in uncontrolled airspace requires effective collision avoidance system. In this study, authors researched the collision avoidance algorithms of small unmanned aerial vehicle based on satisficing theory. In details, we applied the satisficing theory based aircraft conflict resolution method to small unmanned aerial vehicle by considering characteristics of small unmanned aerial vehicles. A simulation environment was constructed and choke point models and performance measures were introduce to evaluate the robustness of proposed methodology. Finally, the relationship between the parameters of the methodology was examined through sensitivity analysis, and the combination of the parameters suitable for operating small unmanned aerial vehicle was searched. This study suggests that the small unmanned aerial vehicles require high rotation angle, small minimum separation and high level of cooperation in the satisficing framework, and operation over 64 agents is not appropriate in the proposed choke point model.