Zhou Kai, Wang Kai, Wang Yuhao, Qu Xiaobo
School of Vehicle and Mobility, Tsinghua University, Beijing, China.
Risk Anal. 2025 Jul;45(7):1698-1715. doi: 10.1111/risa.17685. Epub 2024 Dec 22.
Multifarious applications of unmanned aerial vehicles (UAVs) are thriving in extensive fields and facilitating our lives. However, the potential third-party risks (TPRs) on the ground are neglected by developers and companies, which limits large-scale commercialization. Risk assessment is an efficacious method for mitigating TPRs before undertaking flight tasks. This article incorporates the probability of UAV crashing into the TPR assessment model and employs an A* path-planning algorithm to optimize the trade-off between operational TPR cost and economic cost, thereby maximizing overall benefits. Experiments demonstrate the algorithm outperforms both the best-first-search algorithm and Dijkstra's algorithm. In comparison with the path with the least distance, initially, the trade-off results in a increase in distance while achieving an reduction in TPR. As the trade-off progresses, this relationship shifts, leading to a reduction in the distance with only a negligible increase in TPR by 0.0001, matching the TPR-cost-based algorithm. Furthermore, we conduct simulations on the configuration of UAV path networks in five major cities in China based on real-world travel data and building data. Results reveal that the networks consist of one-way paths that are staggered in height. Moreover, in coastal cities particularly, the networks tend to extend over the sea, where the TPR cost is trivial.
无人机（UAV）的多种应用正在广泛领域蓬勃发展，并便利着我们的生活。然而，地面上潜在的第三方风险（TPR）被开发者和公司忽视了，这限制了大规模商业化。风险评估是在执行飞行任务前减轻第三方风险的有效方法。本文将无人机坠毁的概率纳入第三方风险评估模型，并采用A*路径规划算法来优化运营第三方风险成本与经济成本之间的权衡，从而使整体效益最大化。实验表明，该算法优于最佳优先搜索算法和迪杰斯特拉算法。与距离最短的路径相比，最初，这种权衡导致距离增加，同时实现第三方风险的降低。随着权衡的进行，这种关系发生变化，导致距离减少，而第三方风险仅微增0.0001，与基于第三方风险成本的算法相匹配。此外，我们基于真实的出行数据和建筑数据，对中国五个主要城市的无人机路径网络配置进行了模拟。结果显示，这些网络由高度交错的单向路径组成。而且，特别是在沿海城市，网络倾向于延伸到海上，在那里第三方风险成本微不足道。
