摘要
随着城市超低空物流运输场景的迅速发展,无人机路径规划的安全性显得尤为重要,针对现有路径规划算法无法满足超低空物流运输无人机在密集障碍物场景下进行安全轨迹规划的问题,基于A*算法,将三维环境依据飞行高度划分为多个高度层,以规划风险最小轨迹为目标,从时间、风险两个维度对A*算法的成本估计函数进行重构,从而提出面向城市超低空物流场景的最小危险路径规划算法。仿真实验一表明,本文提出的最小风险路径规划算法在3种不同城市场景,15种不同运行环境中,相比于传统轨迹规划算法,规划得到的路径安全距离平均增加60%,安全性显著提高;将该算法应用于多无人机多高度层的复杂城市场景中,实验结果表明,本文提出的最小风险路径规划算法在兼顾航程的同时可以为多架无人机规划安全性更高的路径;在实验三中运用蒙特卡洛法证明本文算法在路径规划算法中的可靠性与鲁棒性,为城市超低空物流场景提供了安全性更高的路径规划方法。
With the rapid development of urban very-low-level logistics transportation scenarios,the safety of UAV path planning is particularly important.The existing path planning algorithms cannot meet the safety requirements of ultra-low-altitude logistics transportation UAVs.Based on the A*algorithm,the research divides the three-dimensional environment into multiple flight level layers according to the flight height,and aims to plan the minimum risk trajectory by rethinking the cost estimation function from two aspects:time and risk.In order to minimize the risk,a minimum-risk path planning algorithm was proposed for urban very-low-level air transportation scenario.The experiment results show that the minimum-risk path planning algorithm can give 60%higher safe distance on average compared to the traditional trajectory planning algorithms in three different urban scenarios and 15 different operating environments.Also,an additional experiment was conducted to determine the reliability and robustness of this algorithm in a complex urban scenario with multiple UAVs and multiple flight level layers.
作者
程洁
郑远
李诚龙
江波
刘东来
CHENG Jie;ZHENG Yuan;LI Cheng-long;JIANG Bo;LIU Dong-lai(College of Air Traffic Management,Civil Aviation Flight University of China,Guanghan 61830,China;College of Computer Science and Technology,Civil Aviation Flight University of China,Guanghan 618307,China;School of Electronic Information Engineering,Beihang University,Beijing 100191,China)
出处
《科学技术与工程》
北大核心
2023年第2期690-698,共9页
Science Technology and Engineering
基金
国家自然科学基金民航联合基金(U1733105)
四川省中央引导地方科技发展专项(2020ZYD094)
四川省科技计划(2021YFS0391)
民航教育人才类项目(0252103)。
关键词
A*算法
高度层
路径规划
权重系数分配
风险等级
A*algorithm
flight layer
path planning
weight coefficient allocation
risk level
