基于Dubins-RRT算法的无人机集群密集障碍物区域覆盖航路规划  

Dense-obstacle area coverage path planning for UAV swarms based on Dubins-RRT algorithm

作　　者：杨毅 冯彦辉 梁月乾 Yang Yi;Feng Yanhui;Liang Yueqian(Information Sciences Academy,China Electronics Technology Group Corporation,Beijing 100041,China;Unit 93184 of the PLA,Beijing 100076,China)

机构地区：[1]中国电子科技集团公司信息科学研究院,北京100041 [2]中国人民解放军93184部队,北京100076

出　　处：《战术导弹技术》2024年第5期83-98,共16页Tactical Missile Technology

基　　金：科技部科技创新2030-“新一代人工智能”重大项目(2018AAA0102400);国家自然科学基金(U19B2029,U19B2033,U20B2071,U20B2073)。

摘　　要：针对利用无人机集群对城市环境中的指定区域进行协同覆盖的问题,提出了一种两阶段方法。在第一阶段,暂时不考虑建筑物,将待覆盖区域按多个扫描角度进行长方形区域分割,得到最短时间和最小数量两种最优指标下的无避障航路。在第二阶段,在快速搜索随机树(RapidlyExploring Random Tree,RRT)算法的框架下,对采样规则和终止规则进行了重新设计,并直接利用Dubins曲线搜索出相应的无碰撞航路。四组有效性仿真实验以及与扰动流体动态系统方法的比较实验,对所提出方法的有效性和先进性进行了验证。For the cooperative area coverage problem in urban environment by UAV swarm,a two-stage method is proposed.In the first stage,regardless of the obstacles temporarily,the area to be covered is divided into rectangular subareas according to multiple scanning directions,and the non-avoidance path under two performance indices(shortest time and minimum number)is obtained.In the second stage,under the framework of Rapidly-Exploring Random Tree(RRT)algorithm,the sampling rules and termination rules are redesigned,and the corresponding collision-free paths are searched directly using Dubins curve.The effectiveness and advancement of the proposed approach is verified through four groups of effectiveness simulation experiments and comparison experiments with the disturbed fluid dynamics system method.

关 键 词：区域覆盖 航路规划 无人机集群 密集障碍物 Dubins曲线 RRT算法 区域分割 城市环境 

分 类 号：V249[航空宇航科学与技术—飞行器设计]