UAVs’ autonomous collision avoidance in urban space*

Abstract

Small UAVs are seeking wide usage in urban space for its advantageous performance. However, there are crowded with static and dynamic buildings, causing serious challenges for UAVs’ safety. This paper proposes a novel autonomous collision avoidance method based on time-obstacle dynamic map. First, the state estimation and trajectory prediction are performed based on extended Kalman filtering. Second, the time-obstacle dynamic map is constructed via introducing time axis. Third, the flyable paths are searched on the basis of breadth first approach and then the optimal path can be obtained through A-Star algorithm. Finally, the simulation results have shown that the proposed collision avoidance method can avoid immobile building and moving obstacles, and making a safe path for UAVs in urban space