| 33 | 0 | 75 |
| 下载次数 | 被引频次 | 阅读次数 |
针对A*算法在无人机路径规划中平滑度不足、与障碍物距离过近等问题，提出一种梯度下降与A*算法融合的无人机路径规划算法——融合梯度下降的A*算法。基于26邻域搜索的A*算法，在三维环境中生成初始路径；引入具有较强通用性和扩展性的梯度下降优化算法，并结合无人机路径相关约束，构建路径保持、远离障碍物和平滑性三个目标的复合函数，对路径节点进行迭代优化，实现路径的整体平滑与安全提升；在三维复杂城市环境中进行仿真实验。结果表明，融合梯度下降的A*算法相较于原始的A*算法，路径长度减少3.64%，与障碍物最小距离增加69.49%，路径平滑度综合评分提高54.56%。该算法有效提升了无人机在复杂环境中的飞行安全性与路径平滑性，为无人机在复杂环境中路径规划领域提供了技术支撑。
Abstract：To address the issues of insufficient smoothness and excessively close distances to obstacles in UAV path planning with the A~*algorithm,a UAV path planning algorithm that integrates gradient descent with A*,the Gradient Descent-Enhanced A~*Algorithm,has been proposed.Firstly,an initial path is generated in a three-dimensional environment using the A~*algorithm based on 26 neighborhood search.Secondly,a highly generalizable and extensible gradient descent optimization algorithm is introduced,and combined with UAV path-related constraints,a composite function with three objectives path maintenance,obstacle avoidance and smoothness is constructed to iteratively optimize the path nodes,achieving overall path smoothness and safety improvement.Finally,simulation experiments are conducted in a complex three-dimensional urban environment.The results indicate that,compared with the original A~*algorithm,the Gradient Descent-Enhanced A~*Algorithm reduces path length by 3.64%,increases the minimum distance to obstacles by 69.49%,and improves the comprehensive path smoothness score by54.56%,effectively enhancing flight safety and path smoothness for UAVs in complex environments.
[1]陈林，缪志强，王祥科，等．自主飞行器技术及其在低空经济中的应用综述[J]．机器人，2025,47(3):470-496.
[2]徐建新，孙纬，马超．基于改进粒子群算法的无人机三维路径规划[J]．电光与控制，2023,30(6):15-21.
[3]刘文倩，单梁，张伟龙，等．复杂环境下基于改进Informed RRT的无人机路径规划算法[J]．上海交通大学学报，2024,58(4):511-524.
[4]LI Y T,XIE T R,ZHOU J W,et al.Development of unity3D-Based intelligent warehouse visualization platform with enhanced A-Star path planning algorithm[J].Applied sciences,2025,15(22):12202.
[5]陈丽芳，杨火根，陈智超，等．B样条技术与遗传算法融合的全局路径规划[J]．浙江大学学报（工学版），2024,58(12):2520-2530.
[6]郭聚刚，于军琪，冯春勇，等．基于改进A*算法的机器人不平坦地形全局路径规划[J]．计算机工程与应用，2025,61(5):309-322.
[7]FU X,HUANG Z,ZHANG G,et al.Research on path planning of mobile robots based on improved A*algorithm[J].PeerJ.computer science,2025,11:e2691.
[8]LI C G,HUANG X,DING J,et al.Global path planning based on a bidirectional alternating search A*algorithm for mobile robots[J].Computers&industrial engineering,2022,168:108123.
[9]李平伟，张剑高，张财，等．基于改进A*算法的低空多无人机路径规划方法研究[J]．交通科学与工程，2025,41(4):61-68.
[10]祝文杰，李维，王子炎．改进A*算法的无人机城市低空物流路径规划[J]．计算机工程与应用，2025,61(19):302-310.
[11]屈景怡，黄达权，许楠，等．结合A*与速度障碍法的无人机路径规划混合算法[J]．兵器装备工程学报，2025,46(8):70-79.
[12]杨帆，李玮，严天宏．基于梯度优化的移动机器人路径规划算法[J]．现代电子技术，2023,46(21):99-104.
[13]HEUMOS S,GUARRACINO A,SCHMELZLE M N J,et al.Pangenome graph layout by pathguided stochastic gradient descent[J].Bioinformatics,2024,40(7):btae363.
[14]QU J,GU Y,QIU Z,et al.Development of an orchard inspection robot:a ROS-based LiDARSLAM system with hybrid A*-DWA navigation[J].Sensors,2025,25(21):6662.
[15]巩慧，倪翠，王朋，等．基于Dijkstra算法的平滑路径规划方法[J]．北京航空航天大学学报，2024,50(2):535-541.
[16]KHAN Z,ALMOGBIL S,BABAR M,et al.Performance evaluation of pathfinding algorithms for intelligent routing in high-mobility IOT edge networks[J].Procedia computer science,2025,270:182-191.
[17]高九州，张焯．基于改进A*算法的无人机三维空间避障路径规划[J]．计算机测量与控制，2023,31(12):203-209.
[18]DUTT M,BISWAS A,NAGY B.Counting of shortest paths in cubic Grid[J].Acta Polytechnica Hungarica,2024,21(6):169-188.
[19]李杰．基于改进A-Star算法的移动机器人动态路径规划[J]．黑龙江科学，2023,14(16):47-52.
[20]ZHANG Y W,ZHU N Y,ZHANG X Q.A single-mode quasi riemannian gradient descent algorithm for low-multilinear-rank tensor recovery[J].Journal of scientific computing,2026,106:23.
[21]YANG X Y,ZHAO S W,GAO W,et al.Three-dimensional path planning for UAV based on multi-strategy dream optimization algorithm[J].Biomimetics,2025,10(8):551.
[22]张彪，康峰，许舒婷．基于改进RRT与GA的多目标路径规划：以无人机林区巡检为例[J]．北京林业大学学报，2025,47(4):129-141.
[23]顾同成，徐东伟，孙成巨．无人驾驶深度强化学习决策模型性能评测方法综述[J]．计算机工程与应用，2025,61(19):12-42．
基本信息:
DOI：10.13486/j.issn.2097-4973.2026.02.006
中图分类号:V279;V249
引用信息:
[1]高耀辉,王依萍,冯伟伟.基于梯度下降与A~*算法融合的无人机路径规划研究[J].山东航空学院学报,2026,43(02):47-56.DOI:10.13486/j.issn.2097-4973.2026.02.006.
基金信息:
山东省重点研发计划项目(2019GSF109105); 山东航空学院硕士研究生创新基金项目(SHSYCX19)
2026-04-15
2026-04-15