摘要
针对低空经济背景下无人机在复杂三维建筑环境中的路径规划需求,提出改进的双向快速搜索树自适应交替双目标偏差搜索(Sampling-Tree Based bidirectional Rapidly-exploring Random Tree,ST-BA-RRT)算法。该算法在采样阶段采用三维环境下的椭球采样,并配合双目标偏差策略抑制随机树向障碍区扩展,定向引导其向目标生长;扩展阶段运用自适应交替探索与改进人工势场辅助策略,增强算法环境适应性与局部避障能力。碰撞检测环节通过设计新型代价函数减少障碍物检查频次,优化规划时间;连通性处理利用双向随机采样提升规划效率;最后借助β样条平滑路径。实验结果表明,相较于现有算法,ST-BA-RRT算法生成的路径更短、更平滑,路径规划时间平均减少35%,在路径质量与环境适应性方面优势显著,能够高效生成优化飞行轨迹,满足复杂三维建筑环境下无人机路径规划需求。
The path planning requirements of unmanned aerial vehicles(UAVs)in complex three-dimensional building environments under the background of the low-altitude economy are studied.An improved sampling-Tree based bidirectional rapidly-exploring random tree algorithm(ST-BA-RRT)is proposed.During the sampling stage,the proposed algorithm uses the ellipsoidal sampling in a three-dimensional environment,suppresses the expansion of the random tree into the obstacle area in conjunction with the dual-target bias strategy,and guides it to grow towards the target directionally.During the expansion stage,the adaptive alternating exploration and improved artificial potential field auxiliary strategies are applied to enhance the algorithm's environmental adaptability and local obstacle avoidance ability.During the collision detection stage,a new cost function is designed to reduce the frequency of obstacle inspections and optimize the planning time.For the connectivity processing,the bidirectional random sampling is used to improve the planning efficiency.Finally,theβ-spline function is used to smooth the path.The experimental results show that the path generated by the ST-BA-RRT algorithm is shorter and smoother than those generated by the existing algorithms,and the average path planning time is reduced by 35%.The proposed algorithm has significant advantages in terms of path quality and environmental adaptability,efficiently generate the optimized flight trajectories,and meets the path planning requirements of UAVs in complex three-dimensional building environments.
作者
郑振岗
李新凯
孟月
张宏立
ZHENG Zhengang;LI Xinkai;MENG Yue;ZHANG Hongli(School of Intelligent Science and Technology,Xinjiang University,Urumqi 830017,Xinjiang,China)
出处
《兵工学报》
北大核心
2026年第1期181-199,共19页
Acta Armamentarii
基金
国家自然科学基金项目(62263030)
新疆维吾尔自治区自然科学基金项目(2023D01C187)
新疆维吾尔自治区“天池英才”引进计划项目(5105240151d)
“天山英才”科技创新领军人才项目(2024TSYCLJ0008)。
关键词
无人机
改进的双向快速搜索树
椭球化采样
双目标偏差策略
自适应交替探索
unmanned aerial vehicles
improved bidirectional rapidly-exploring random tree
ellipsoidal sampling
dual-target bias strategy
adaptive alternating exploration
