面向非结构障碍物避障的A^(*)算法改进  

UAV A^(*)Algorithm Improvement for Obstacle Avoidance of Non-structured Obstacles

作　　者：周恒旭 熊治国 肖晓波 ZHOU Heng-xu;XIONG Zhi-guo;XIAO Xiao-bo(School of Aviation,Beijing Institute of Technology,Zhuhai 519000 China)

机构地区：[1]北京理工大学珠海学院航空学院,广东珠海519000

出　　处：《自动化技术与应用》2025年第2期9-12,84,共5页Techniques of Automation and Applications

基　　金：广东省普通高校特色创新项目(2020KTSCX190)。

摘　　要：随着无人自主系统在各个领域应用的深入,对自主避障算法提出越来越高的要求。针对无人自主系统面向非结构障碍物避障的问题,从启发函数h(n)的选择和路径修剪策略的设计这两个方面上改进了经典的A*算法,使其生成的路径拐点尽可能少。最后基于Gazebo和ROS进行无人机仿真验证,结果表明改进的A^(*)算法在面对非结构障碍物场景时的避障效果良好。As the application of unmanned autonomous systems in various fields deepens,higher and higher requirements are put forward for autonomous obstacle avoidance algorithms.To address the problem of unmanned autonomous system obstacle avoidance for unstructured obstacles,this paper improves the classical A^(*)algorithm from two aspects,namely the selection of heuristic function h(n)and the design of path pruning strategy,so that it generates as few path inflection points as possible.Finally,the UAV simula-tion is verified based on Gazebo and ROS,and the results show that the improved A*algorithm is effective in obstacle avoidance in the face of non-structural obstacle scenarios.

关 键 词：非结构障碍物 双目摄像头 A^(*)算法 仿真验证 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] V279[自动化与计算机技术—控制科学与工程]