融合图神经网络的移动机器人集群轨迹规划算法

doi:10.13229/j.cnki.jdxbgxb.20250415

摘要/Abstract

摘要：

针对移动机器人集群作业时，因存在路径冲突导致机器人易出现碰撞，研究了融合图神经网络的移动机器人集群轨迹规划算法。构建了移动机器人运动学模型，描述了机器人运动特性；运用基于融合图神经网络的移动机器人集群运行环境特征提取模型，提取移动机器人集群运行环境多维度特征，全面了解环境信息；结合人工势场法，依据所提取特征，将目的地设成引力源，吸引机器人的运动方向为目的地方向；将障碍物设为斥力源，控制机器人避开障碍物，每个机器人受合力决定运动方向与速度，规划移动机器人集群运行时防碰撞轨迹。实验显示：所提算法可有效提取移动机器人集群运行环境多维特征，为多个移动机器人规划无碰撞、风险指数低于0.1的运行轨迹。

关键词: 融合图神经网络, 残差连接, 移动机器人, 集群轨迹规划, 人工势场法, 防碰撞

Abstract:

Aiming at the problem of path conflicts among multiple robots running in the same environment during mobile robot cluster operations， which can lead to collisions among robots， this paper studies a trajectory planning algorithm for mobile robot clusters that integrates graph neural networks. Construct a kinematic model of a mobile robot to describe its motion characteristics； Using a fusion graph neural network-based mobile robot cluster operating environment feature extraction model， multi-dimensional features of the mobile robot cluster operating environment are extracted to comprehensively understand environmental information； Combining the artificial potential field method， based on the extracted features， the destination is set as a gravitational source， and the direction of motion that attracts the robot is towards the destination； Set obstacles as repulsive sources， control robots to avoid obstacles， and determine the direction and speed of each robot's movement based on the resultant force. Plan collision avoidance trajectories for mobile robot clusters during operation. The experiment shows that the proposed algorithm can effectively extract multidimensional features of the operating environment of mobile robot clusters， and plan collision free and risk index less than 0.1 running trajectories for multiple mobile robots.

Key words: fusion graph neural network, residual connection, mobile robots, cluster trajectory planning, artificial potential field method, anti-collision

中图分类号:

TP391

顾炜江,业巧林,王兴虎. 融合图神经网络的移动机器人集群轨迹规划算法[J]. 吉林大学学报(工学版), 2026, 56(7): 2034-2040.

Wei-jiang GU,Qiao-lin YE,Xing-hu WANG. Trajectory planning algorithm for mobile robot clusters based on fused graph neural networks[J]. Journal of Jilin University(Engineering and Technology Edition), 2026, 56(7): 2034-2040.

图/表 6

图1

图2

表1

图3

图4

图5

参考文献 15

[1]	武星, 汤凯, 李兴达, 等. 多视角雷达点云融合的移动机器人集群跟踪[J]. 仪器仪表学报, 2023, 44(12): 175-186.
	Wu Xing, Tang Kai, Li Xing-da, et al. Mobile robot cluster tracking based on multi-perspective Lidar point cloud fusion[J]. Chinese Journal of Scientific Instrument, 2023,44(12):175-186.
[2]	陈晨, 殷诗雨, 邱星星, 等. 采样数据交互的异构非完整轮式机器人集群连通保持切换编队控制[J]. 南京理工大学学报: 自然科学版,2024,48(5):615-625, 634.
	Chen Chen, Yin Shi-yu, Qiu Xing-xing, et al. Connectivity-preserving switching formation control of heterogeneous nonholonomic wheeled robots with sampled-data interaction[J]. Journal of Nanjing University of Science and Technology, 2024, 48(5):615-625, 634.
[3]	郭万金, 李儒, 郝钦磊, 等. 农业物料移运机器人协同作业时间最优轨迹规划方法[J]. 农业机械学报,2024, 55(1): 22-38.
	Guo Wan-jin, Li Ru, Hao qin-lei, et al. Time-optimal trajectory planning method for cooperative working of agriculture material handling robot[J]. Transactions of the Chinese Society for Agricultural Machinery, 2024,55(1):22-38.
[4]	杨振, 李俊丽, 杨立炜, 等. 安全性A^*融合DWA的分布式多移动机器人路径规划方法[J]. 控制工程,2024, 31(12): 2284-2295.
	Yang Zhen, Li Jun-li, Yang Li-wei, et al. Distributed multi-mobile robots path planning method based on safe A^*fused with dynamic window approach[J]. Control Engineering of China, 2024,31(12):2284-2295.
[5]	张凯翔, 毛剑琳, 向凤红, 等. 基于讨价还价博弈机制的B-IHCA*多机器人路径规划算法[J]. 自动化学报, 2023, 49(7): 1483-1497.
	Zhang Kai-xiang, Mao Jian-lin, Xiang Feng-xiang, et al. B-IHCA*,a bargaining game based multi-agent path finding algorithm[J]. Acta Automatica Sinica, 2023,49(7):1483-1497.
[6]	张凯翔, 毛剑琳, 宣志玮, 等. 面向关隘地形的分层调度多机器人路径规划[J]. 计算机集成制造系统,2024, 30(1): 172-183.
	Zhang Kai-xiang, Mao Jian-lin, Xuan Zhi-wei, et al. Hierarchical scheduling based multi-robot path planning for pass terrain[J]. Computer Integrated Manufacturing Systems, 2024,30(1):172-183.
[7]	褚晶, 周力, 岳颀, 等. 一种基于Petri网的多机器人路径规划建模方法[J]. 西北工业大学学报, 2024, 42(4): 716-725.
	Chu Jing, Zhou Li, Yue Qi, et al. A Petri nets-based modeling method for multi robot path planning[J]. Journal of Northwestern Polytechnical University, 2024,42(4):716-725.
[8]	毛剑琳, 贺志刚, 张书凡, 等. 面向多到一任务交接的多机器人协作路径规划算法[J]. 仪器仪表学报,2024, 45(9): 237-248.
	Mao Jian-lin, He Zhi-gang, Zhang Shu-fan, et al. Multi-robot collaborative path planning algorithm for many-to-one task handover[J]. Chinese Journal of Scientific Instrument, 2024,45(9):237-248.
[9]	程晓明, 刘银华, 赵文政. 面向大型三维结构检测的多机器人覆盖路径规划方法[J]. 计算机集成制造系统, 2023, 29(1): 246-253.
	Cheng Xiao-ming, Liu Yin-hua, Zhao Wen-zheng. Multi-robot coverage path planning for large 3D structure inspection[J]. Computer Integrated Manufacturing Systems, 2023,29(1):246-253.
[10]	Lewke M, Wu H J, List A, et al.Automated trajectory planning and analytical improvement for automated repair by robot-guided cold spray[J].Journal of Thermal Spray Technology,2024,33(2/3):515-529.
[11]	陈金杰, 王一蕾, 傅仰耿. 注意力融合机制和拓扑关系挖掘的异构图神经网络[J]. 福州大学学报: 自然科学版, 2025, 53(1): 1-9.
	Chen Jin-jie, Wang Yi-lei, Fu Yang-geng. Heterogeneous graph neural network based on attention fused mechanisms and topology relation mining[J]. Journal of Fuzhou University (Natural Science Edition), 2025,53(1):1-9.
[12]	秦志龙, 朱一峰, 邓琨, 等. 融合路径优化的异构图神经网络算法[J]. 小型微型计算机系统, 2025, 46(3): 627-635.
	Qin Zhi-long, Zhu Yi-feng, Deng Kun, et al. Path Optimization based heterogeneous graph neural network algorithm[J]. Journal of Chinese Computer Systems, 2025, 46(3): 627-635.
[13]	刘冠麟, 赵志耘, 曾文. 一种基于图神经网络的技术融合预测方法[J]. 情报杂志, 2024, 43(12): 117-124, 185.
	Liu Guan-lin, Zhao Zhi-yun, Zeng Wen. A technology convergence prediction method based on graph neural networks[J]. Journal of Intelligence, 2024, 43(12):117-124, 185.
[14]	江昌旭, 卢玥君, 邵振国, 等. 基于图神经网络多智能体强化学习的电力-交通融合网协同优化运行[J]. 高电压技术, 2023, 49(11): 4622-4631.
	Jiang Chang-xu, Lu Yue-jun, Shao Zhen-guo, et al. Collaborative optimization operation of integrated electric power and traffic network based on graph neural network multi-agent reinforcement learning[J]. High Voltage Engineering, 2023, 49(11): 4622-4631.
[15]	旺扎拉, 戴景民. 无冲突协调自动联控集群机器人方法与仿真[J]. 计算机仿真, 2024, 41(4): 451-455.
	Wang Zha-la, Dai Jing-min. Simulation of conflict free coordination and automatic joint control method for cluster robots[J]. Computer Simulation, 2024, 41(4): 451-455.

相关文章 15

[1]	张伏,韩伟东,鲍若飞,张亚坤,王亚飞,付三玲. 融合改进A^*与DWA算法的车间移动机器人路径规划[J]. 吉林大学学报(工学版), 2025, 55(9): 3020-3031.
[2]	曹毅,夏宇,高清源,叶培涛,叶凡. 基于超连接图卷积网络的骨架行为识别方法[J]. 吉林大学学报(工学版), 2025, 55(2): 731-740.
[3]	聂建军,侯军凯,解晓琳,鄢鸿桢. 新型巡检机器人移动底盘设计及越障性能分析[J]. 吉林大学学报(工学版), 2024, 54(2): 356-364.
[4]	吴振宇,刘小飞,王义普. 基于DKRRT^*-APF算法的无人系统轨迹规划[J]. 吉林大学学报(工学版), 2023, 53(3): 781-791.
[5]	祁贤雨,王巍,王琳,赵玉飞,董彦鹏. 基于物体语义栅格地图的语义拓扑地图构建方法[J]. 吉林大学学报(工学版), 2023, 53(2): 569-575.
[6]	王秀芳,孙双,丁春阳. 基于1D⁃RSCNN的嵌入式轴承故障实时检测[J]. 吉林大学学报(工学版), 2022, 52(2): 310-317.
[7]	赵宏伟,张子健,李蛟,张媛,胡黄水,臧雪柏. 基于查询树的双向分段防碰撞算法[J]. 吉林大学学报(工学版), 2021, 51(5): 1830-1837.
[8]	刘明,荣学文,李贻斌,张帅帅,尹燕芳,阮久宏. 基于地形聚类分析的移动机器人速度自适应控制[J]. 吉林大学学报(工学版), 2021, 51(4): 1496-1505.
[9]	于树友,常欢,孟凌宇,郭洋,曲婷. 基于扰动观测器的轮式移动机器人滚动时域路径跟踪控制[J]. 吉林大学学报(工学版), 2021, 51(3): 1097-1105.
[10]	郭继昌,吴洁,郭春乐,朱明辉. 基于残差连接卷积神经网络的图像超分辨率重构[J]. 吉林大学学报(工学版), 2019, 49(5): 1726-1734.
[11]	顾万里,王萍,胡云峰,蔡硕,陈虹. 具有H_∞性能的轮式移动机器人非线性控制器设计[J]. 吉林大学学报(工学版), 2018, 48(6): 1811-1819.
[12]	王鑫, 贾庆轩, 高欣, 赵兵, 崔宝江. 高效无线射频识别自适应型跟踪树防碰撞算法[J]. 吉林大学学报(工学版), 2015, 45(4): 1225-1233.
[13]	叶锦华,李迪,叶峰. 轮式移动机器人的双强化学习自适应模糊控制[J]. 吉林大学学报(工学版), 2014, 44(3): 742-749.
[14]	苑全德, 关毅, 洪炳镕, 孟祥萍. 一种异质多移动机器人系统及其基于MAS的协作机制[J]. 吉林大学学报(工学版), 2013, 43(01): 141-146.
[15]	贾鹤鸣, 宋文龙, 陈子印. 基于反馈增益反步法的非完整约束移动机器人路径跟踪控制[J]. , 2012, 42(04): 1003-1007.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

参数类型	具体信息
额定载重/kg	150
尺寸/mm	805705605
巡点精度/mm	±5.5
导引方式	RFID磁带导引
充电方式	电源
电源	49/65 A·h铅酸蓄电池
驱动模式	两轮差速