低传感受限空间虚拟管道下多无人机轨迹跟踪与去冲突协同控制

Cooperative control for multi-UAV trajectory tracking and deconfliction in low-sensing confined spaces with virtual tubes

摘要

针对工业厂房、综合管廊与仓储园区等受限空间中多无人机（UAVs）协同巡检在通道狭窄、遮挡频发条件下易发生拥塞与近距交互冲突，且定位不稳定与机载算力有限的问题，本文提出一种基于虚拟管道几何分层与学习增强的多无人机协同去冲突方法. 在几何规划层构建中心线与变半径的虚拟管道，并在弗勒内标架下生成可分配子通道与连续参考子轨迹，实现宏观空间隔离；在学习层，设计冲突感知共享参数多智能体近端策略优化控制器，构建由进度、偏差、安全裕度与邻机信息等组成的低维结构化观测，将决策约束为切向速度意图，并由纯追踪方法提供横向几何纠偏，在机间距离小于安全阈值时叠加势场型排斥速度项以抑制近距冲突. 基于 AirSim 仿真平台的 5 机迷宫场景实验显示，飞行过程未发生碰撞，机间最小距离始终大于 0.6 m；在相同任务完成率条件下，相较分布式规则控制器，平均横向跟踪误差由 0.571 m 降低至 0.201 m，降幅为 64.8%. 此外，在未见过的非一致性复杂场景中，所提方法实现了零样本迁移并保持零碰撞，验证了策略在未知几何约束下良好的结构泛化能力. 方法可在低传感约束下兼顾协同通行安全与轨迹跟踪精度.

Abstract

The cooperative inspection of multiple unmanned aerial vehicles (UAVs) in confined spaces, such as industrial plants, utility tunnels, and warehouse parks, presents significant challenges. These environments often feature narrow passages, frequent occlusions, and complicated layouts, resulting in congestion and close-range interaction conflicts among UAVs. Additionally, the instability of localization systems and the limited computational power available on UAVs further hinder their effective deployment in such environments. These conditions require novel solutions for safe and efficient UAV operation. To address these challenges, in this study, a cooperative deconfliction method, which integrates virtual-tube-based geometric layering with learning-enhanced control mechanisms, is proposed. In the geometric planning layer, a virtual tube is constructed using a parameterized centerline and variable radius, enabling the creation of allocable sub-channels. Continuous reference sub-trajectories are generated within the Frenet frame to ensure macroscopic spatial separation of the UAVs. This approach ensures that UAVs are kept within safe operational zones and prevents collisions using geometry-based constraints. Furthermore, in the learning-based control layer, a conflict-aware multi-agent proximal policy optimization (MA-PPO) controller with shared parameters is developed. This controller utilizes low-dimensional structured observations, which include task progress, lateral deviation, safety margin, and neighboring UAV information. The decision-making process is constrained to outputting a tangential-speed intention, which reduces the complexity of the control problem and ensures easier training. A pure pursuit method provides lateral geometric correction. When the inter-agent distance falls below a safety threshold, a potential-field-based repulsive velocity term is added to suppress close-range conflicts. To validate the proposed method, simulation experiments were conducted using a five-UAV maze scenario on the AirSim simulation platform. The experiments demonstrate that the UAVs successfully complete the mission without any collisions, with the minimum inter-agent distance consistently remaining above 0.6 m. Additionally, the average lateral tracking error decreased from 0.571 m to 0.201 m, representing a reduction of 64.8% when compared to a distributed rule-based control method. These results showcase the efficacy of the proposed method in terms of trajectory tracking accuracy and cooperative safety in confined spaces. Moreover, the proposed method was tested in a complex, inconsistent scenario, which was not included in the training environment. The results show that the method achieved zero-sample transfer and maintained zero collisions, proving the robustness and generalization ability of the learned policy under unknown geometric constraints. This demonstrates that the method is not only effective in the training environment but also capable of adapting to new, unforeseen scenarios. Overall, the results indicate that the proposed approach can simultaneously ensure safe cooperative passage and accurate trajectory tracking under low-sensing constraints. The integration of geometric constraints with learning-based decision-making provides a promising solution for multi-UAV cooperative missions in confined spaces. By addressing the challenges of low sensing and limited computational power, this method paves the way for the practical deployment of UAVs in complex and restricted environments.

关键词

多无人机协同 / 受限空间 / 虚拟管道 / 主动去冲突 / MA-PPO-PS

Key words

multi-UAV coordination / confined-space operations / virtual tube / active deconfliction / MA-PPO-PS

参考文献

[1] Liao X H, Xu C C, Ye H P. Benefits and challenges of constructing low—altitude air route network infrastructure for developing low—altitude economy[J]. Bull Chin Acad Sci, 2024, 39(11): 1966

[2] （廖小罕, 徐晨晨, 叶虎平. 低空经济发展与低空路网基础设施建设的效益和挑战[J]. 中国科学院院刊, 2024, 39（11）： 1966）

[3] Zhang R, Hao G B, Zhang K, et al. Unmanned aerial vehicle navigation in underground structure inspection: A review[J]. Geol J, 2023, 58(6): 2454

[4] Javed S, Hassan A, Ahmad R, et al. State—of—the—art and future research challenges in UAV swarms[J]. IEEE Internet Things J , 2024, 11(11): 19023

[5] Sun J, Xu G T, Wang Z, et al. Safe flight corridor constrained sequential convex programming for efficient trajectory generation of fixed—wing UAVs[J]. Chin J Aeronaut, 2025, 38(1): 103174

[6] Chang Y X, Cheng Y Q, Manzoor U, et al. A review of UAV autonomous navigation in GPS—denied environments[J]. Rob Auton Syst, 2023, 170: 104533

[7] Mei Y L, Cui L K, Hu X Y, et al. Obstacle avoidance and formation control of multiple unmanned vehicles in complex environments based on artificial potential field method[J]. Chin J Eng, 2025, 47(2): 364

[8] （梅艺林, 崔立堃, 胡雪岩, 等. 基于人工势场法的复杂环境下多无人车避障与编队控制[J]. 工程科学学报, 2025, 47（2）： 364）

[9] Liang C Q, Liu L, Liu C. Multi—UAV autonomous collision avoidance based on PPO—GIC algorithm with CNN—LSTM fusion network[J]. Neural Networks, 2023, 162: 21

[10] Liu Z W, Sun R, Jiang L. Robust adaptive position algorithm for GNSS/IMU based on pseudorange residual and innovation[J]. J Beijing Univ Aeronaut Astronaut, 2024, 50(4): 1316

[11] （刘正午, 孙蕊, 蒋磊. 基于伪距残差和新息的GNSS/IMU抗差自适应定位算法[J]. 北京航空航天大学学报, 2024, 50（4）： 1316）

[12] Han L, Wang Y, Yan Z W, et al. Event—triggered formation control with obstacle avoidance for multi—agent systems applied to multi—UAV formation flying[J]. Control Eng Pract, 2024, 153: 106105

[13] Feng S Y, Zeng L Z, Liu J N, et al. Multi—UAVs collaborative path planning in the cramped environment[J]. IEEE/CAA J Autom Sin, 2024, 11(2): 529

[14] Vesentini F, Muradore R, Fiorini P. A survey on velocity obstacle paradigm[J]. Rob Auton Syst, 2024, 174: 104645

[15] Xia K W, Li X Y, Li K D, et al. Distributed predefined—time control for cooperative tracking of multiple quadrotor UAVs[J]. IEEE/CAA J Autom Sin, 2024, 11(10): 2179

[16] Chen M, Liu W, Zhang P. Distributed collision avoidance tracking control for quadrotor cooperative suspension system under performance constraints[J]. Acta Autom Sin, 2024, 50(12): 2392

[17] （陈谋, 刘伟, 张鹏. 性能约束下的四旋翼无人机协同吊挂系统分布式避碰跟踪控制[J]. 自动化学报, 2024, 50（12）： 2392）

[18] Chandra R, Zinage V, Bakolas E, et al. Deadlock—free, safe, and decentralized multi—robot navigation in social mini—games via discrete—time control barrier functions[J]. Auton Rob, 2025, 49(2): 12

[19] Li B, Song C, Bai S X, et al. Multi—UAV trajectory planning during cooperative tracking based on a fusion algorithm integrating MPC and standoff[J]. Drones, 2023, 7(3): 196

[20] Mao P D, Lv S L, Quan Q. Tube RRT*: Efficient homotopic path planning for swarm robotics passing—through large—scale obstacle environments[J]. IEEE Rob Autom Lett, 2025, 10(3): 2247

[21] Lin Y, Na Z Y, Feng Z L, et al. Dual—gamebased UAV swarm obstacle avoidance algorithm in multi—narrow type obstacle scenarios[J]. EURASIP J Adv Signal Process, 2023, 2023(1): 118

[22] Hao G Q, Lv Q, Huang Z, et al. UAV path planning based on improved artificial potential field method[J]. Aerospace, 2023, 10(6): 562

[23] Niu H T, Ma C B, Han P. Directional optimal reciprocal collision avoidance[J]. Rob Auton Syst, 2021, 136: 103705

[24] Wang C, Wei C S, Yin Z Y, et al. Collaborative planning of multi—UAV trajectories and communication strategies considering channel resource constraints[J]. Acta Aeronaut Astronaut Sin, 2025, 46(18): 331837

[25] （王辰, 魏才盛, 殷泽阳, 等. 考虑信道资源约束的多无人机航迹与通信策略协同规划[J]. 航空学报, 2025, 46（18）： 331837）

[26] Hong X T, Wang Z J, Wang Y, et al. Multi—UAV dynamic target search based on multi—potential—field fusion reward shaping MAPPO[J]. Drones, 2025, 9(11): 770

[27] Xu Z F, Han X M, Shen H Y, et al. NavRL: Learning safe flight in dynamic environments[J]. IEEE Rob Autom Lett, 2025, 10(4): 3668

[28] Razzaghi P, Tabrizian A, Guo W, et al. A survey on reinforcement learning in aviation applications[J]. Eng Appl Artif Intell, 2024, 136: 108911

[29] Quan Q, Gao Y, Bai C G. Distributed control for a robotic swarm to pass through a curve virtual tube[J]. Rob Auton Syst , 2023, 162: 104368

[30] Ma Z W, Wang Z M, Ma A T, et al. A low—altitude obstacle avoidance method for UAVs based on polyhedral flight corridor[J]. Drones, 2023, 7(9): 588

[31] Wang M Y, Zhang D, Li C Y, et al. Multiple fixed—wing UAVs collaborative coverage 3D path planning method for complex areas[J]. Def Technol, 2025, 47: 197

[32] Mao P D, Fu R, Quan Q. Optimal virtual tube planning and control for swarm robotics[J]. Int J Rob Res , 2024, 43(5): 602

[33] Rysdyk R. Unmanned aerial vehicle path following for target observation in wind[J]. J Guid Control Dyn, 2006, 29(5): 1092

[34] Hanson A J, Ma H. Parallel transport approach to curve framing[R/OL]. 2025—11—05. https://scholarworks.iu.edu/dspace/items/39fbe931—c2c8—43a1—9c9d—a207b92d551e/full

[35] Bishop R L. There is more than one way to frame a curve[J]. Am Math Mon , 1975, 82(3): 246

[36] Perlin K. Improving noise[J]. ACM Trans Graphics, 2002, 21(3): 681

[37] Xu G T, Wang Z, Cao Y, et al. Dynamic—priority—decoupled UAV swarm trajectory planning using distributed sequential convex programming[J]. Acta Aeronaut Astronaut Sin, 2022, 43(2): 325059

[38] （徐广通, 王祝, 曹严, 等. 动态优先级解耦的无人机集群轨迹分布式序列凸规划[J]. 航空学报, 2022, 43（2）： 325059）

[39] Zhao J, Pei Z N, Jiang B, et al. Virtual tube visual obstacle avoidance for UAV based on deep reinforcement learning[J]. Acta Autom Sin, 2024, 50(11): 2245

[40] （赵静, 裴子楠, 姜斌, 等. 基于深度强化学习的无人机虚拟管道视觉避障[J]. 自动化学报, 2024, 50（11）： 2245）

[41] Crann V, Amiri P, Knox S, et al. Decentralized deconfliction of aerial robots in high intensity traffic structures[J]. J Field Rob, 2024, 41(5): 1541

基金资助

新疆维吾尔自治区天山英才青年拔尖人才资助项目(2024TSYCCX0023)

国家自然科学基金资助项目(62263030)