Drones have been greatly developed to facilitate the progress of various industries. The safe operation of drones in the urban airspace is critical to ensure a reliable and high-efficient urban air traffic system. This work presents a fusion scheme to achieve autonomous drone collision-free path planning considering static obstacles and dynamic threats detected. Firstly, a 3D voxel jump point search (JPS) based path planning model is developed to generate the static collision-free reference path. With the optimization, the reference path is then de-diagonalized, reconstructed, and smoothed to obtain the desired path. Subsequently, a local collision resolution method is proposed to avoid near mid-air collision of the dynamic threats. The method is based on the Markov decision process (MDP) to implement real-time dynamic collision avoidance. Simulations are conducted to verify the performance of the proposed model. The simulation results demonstrate that the proposed model is effective to achieve the autonomous path planning and real-time collision resolution of multirotor drones.

无人机得到了极大的发展，以促进各个行业的进步。无人机在城市空域的安全运行对于确保可靠、高效的城市空中交通系统至关重要。这项工作提出了一种融合方案，以实现考虑静态障碍物和检测到的动态威胁的自主无人机无碰撞路径规划。首先，开发了基于 3D 体素跳跃点搜索 (JPS) 的路径规划模型来生成静态无碰撞参考路径。通过优化，参考路径然后被去对角化、重建和平滑以获得期望的路径。随后，提出了一种局部碰撞解决方法来避免动态威胁的近空中碰撞。该方法基于马尔可夫决策过程（MDP）来实现实时动态避碰。进行仿真以验证所提出模型的性能。仿真结果表明，所提出的模型能够有效地实现多旋翼无人机的自主路径规划和实时碰撞解决。