面向非结构障碍物避障的A^(＊)算法改进.pdfVIP

控制理论与应用《自动化技术与应用》2025年第44卷第2期

ControlTheoryandApplications

DOI:10.20033/j.1003-7241.（2025）02-0009-05.

面向非结构障碍物避障的A*算法改进

周恒旭,熊治国,肖晓波

(北京理工大学珠海学院航空学院,广东珠海519000)

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。

*

关键词:非结构障碍物;双目摄像头;A算法;仿真验证

中图分类号：TP18;V279文献标识码:A文章编号:1003-7241(2025)02-0009-05

UAVA*AlgorithmImprovementforObstacleAvoidanceof

Non-structuredObstacles

ZHOUHeng-xu,XIONGZhi-guo,XIAOXiao-bo

(SchoolofAviation,BeijingInstituteofTechnology,Zhuhai519000China)

Abstract:Astheapplicationofunmannedautonomoussystemsinvariousfieldsdeepens,higherandhigherrequirementsareputforward

forautonomousobstacleavoidancealgorithms.Toaddresstheproblemofunmannedautonomoussystemobstacleavoidancefor

unstructuredobstacles,thispaperimprovestheclassicalA*algorithmfromtwoaspects,namelytheselectionofheuristicfunction

h(n)andthedesignofpathpruningstrategy,sothatitgeneratesasfewpathinflectionpointsaspossible.Finally,theUAVsimula-

tionisverifiedbasedonGazeboandROS,andtheresultsshowthattheimprovedA*algorithmiseffectiveinobstacleavoidance

inthefaceofnon-structuralobstaclescenarios.

Ke