复杂城市环境下无人机路网模型研究

Route Network Modeling for Unmanned Aerial Vehicle in Complex Urban Environment

针对复杂城市环境下无人机路径规划问题,采用三维可视图法研究路网模型。首先,在考虑无人机飞行安全裕度的前提下,将城市密集而不规则的障碍物环境进行变形重组,再以不同的水平和竖直间隔对障碍物外表面进行离散化的节点采集,并构建基于三维可视图的复杂城市低空路网模型。其次,为降低无人机之间的潜在冲突和碰撞风险,引入无人机机动保护区的概念,进一步缩减路网规模,优化路网结构。最后,结合无人机性能和平稳飞行的要求,以最大航向角改变量作为主要限制条件,以最小化路径长度为目标,提出改进的涟漪扩散算法进行求解。仿真结果表明:三维可视图中的采点间隔直接决定了路网模型中节点和链接的数量,并对最优路径与规划时间具有显著影响;1000组仿真实验表明,考虑机动保护区后,最短路径的平均长度相较于无机动保护区时增长了不足1%,而计算耗时降低了近70%。仿真实验验证,通过引入无人机机动保护区和航向角改变量的限制,能够有效降低路网规模,提升运算效率,并有利于获得平滑的路径,降低无人机的潜在碰撞风险。

This paper focuses on the path planning problem of unmanned aerial vehicle(UAV)in complex urban environments and uses the three-dimensional(3D)visibility graph method for the road network modeling.First,the dense and irregular obstacle environment of the city is deformed and reorganized under the premise considering the safety margin of UAV flight,and then the discrete nodes are collected on the outer surface of the obstacles at different horizontal and vertical intervals.A complex urban low-altitude road network model is developed based on 3D visibility graphs.To reduce the potential conflicts and collision risks between UAVs,the concept of UAV maneuvering protection zone is introduced to further reduce the road network scale and optimize the road network structure.Combined with the requirements of UAV performance and smooth flight,an improved ripple spreading algorithm is proposed to solve the problem with the maximum heading angle change as the main constraint and the minimum path length as the goal.The simulation results show that:the interval of picking points in the 3D visibility graph directly determines the number of nodes and links in the road network model and has a significant impact on the optimal path and planning time.1000 sets of simulation experiments show that the average length of the shortest path after considering the maneuvering protection zone increases by less than 1%compared with that without the maneuvering protection zone,while the computation time is reduced by nearly 70%.The simulation experiments verify that the introduction of UAV maneuvering protection zone and the limitation of heading angle change can effectively reduce the size of the road network,improve the computational efficiency,and help to obtain smooth paths and reduce the potential collision risk of UAVs.