历史轨迹驱动的无人机自主着陆迭代学习控制

  • 田秋扬 ,
  • 王泽林 ,
  • 胡天江
展开
  • 中山大学

收稿日期: 2025-07-28

  修回日期: 2025-09-30

  网络出版日期: 2025-10-17

基金资助

国家自然科学基金

History-trajectory-driven iterative learning control for autonomous fixed-wing UAV landing

  • TIAN Qiu-Yang ,
  • WANG Ze-Lin ,
  • HU Tian-Jiang
Expand

Received date: 2025-07-28

  Revised date: 2025-09-30

  Online published: 2025-10-17

摘要

小型固定翼无人机在林业、消防等低空经济领域已广泛应用,其任务执行常需依赖可靠的自主着陆能力。尽管着陆过程中无人机的位姿模态变化具有重复性,但受低空风扰及初始状态随机偏移影响,实际着陆轨迹易出现轨迹长度波动并威胁着陆安全。针对此问题,本文提出一种基于历史轨迹驱动的迭代学习控制方案。该方案在迭代维度上,利用历史轨迹数据优化期望着陆轨迹,间接降低跟踪误差,既兼容现有飞控框架又避免频繁调参;在时间维度上,采用动态时间规整算法对多架次历史轨迹进行非线性对齐,突破了传统迭代学习要求轨迹长度严格一致的约束,并确保了迭代收敛性。仿真验证结果表明,相较于传统自动驾驶仪,该方案使着陆轨迹的横侧向和纵向跟踪误差均降低50%以上,有效提升了自主着陆的成功率与安全性。

本文引用格式

田秋扬 , 王泽林 , 胡天江 . 历史轨迹驱动的无人机自主着陆迭代学习控制[J]. 航空学报, 0 : 1 -0 . DOI: 10.7527/S1000-6893.2025.32634

Abstract

Small fixed-wing Unmanned Aerial Vehicles (UAVs) have been widely adopted in low-altitude economic sectors such as forestry and firefighting, where reliable autonomous landing capabilities are essential for mission execution. Although the positional and atti-tude modal changes of UAVs during landing exhibit repetitive patterns, low-altitude wind disturbances and random initial state devia-tions frequently cause fluctuations in trajectory length, compromising landing safety. To address this issue, this paper proposes a historical-trajectory-driven Iterative Learning Control (ILC) scheme. At the iterative dimension, the scheme optimizes the desired landing trajectory by leveraging historical trajectory data, indirectly reducing tracking errors while maintaining compatibility with existing flight control frameworks and avoiding frequent parameter adjustments. At the temporal dimension, the Dynamic Time Warping (DTW) algorithm is employed to nonlinearly align multi-sortie historical trajectories across time and space. This approach overcomes the traditional iterative learning constraint requiring strictly identical desired trajectories while ensuring iterative conver-gence. Simulation results demonstrate that, compared to conventional autopilots, this scheme reduces lateral and longitudinal tracking errors in landing trajectories by over 50%, significantly enhancing the success rate and safety of autonomous landings.

参考文献

[1]张军, 陈磊, 高智杰.低空无人机技术研究现状与展望[J].中国工程科学, 2025, 27(2):73-85
[2]ZHANG J, CHEN L, GAO Z J, et al.Low-altitude un manned aerial vehicle technology: current status and pro-spects[J].Strategic Study of CAE, 2025, 27(2):73-85
[3]蒲钒, 陈志杰, 朱永文.数字低空融合运行空中交通管理技术[J].航空学报, 2025, 46(11):531331-1-531331-23
[4]PU F, CHEN Z J, ZHU Y W, et al.Air traffic manage-ment technologies for digital low-altitude integrated oper-ations[J].Acta Aeronautica et Astronautica Sinica, 2025, 46(11):531331-1-531331-23
[5]ZHAO R, LI Y, GAO F, et al.Multi-agent constrained policy optimization for conflict-free management of con-nected autonomous vehicles at unsignalized intersections[J].IEEE Transactions on Intelligent Transportation Systems, 2024, 25(6):5374-5388
[6]QI PY, ZHAO X W, Rafael Palacios.Autonomous land-ing control of highly flexible aircraft based on lidar pre-view in the presence of wind turbulence[J].IEEE Trans-actions on Aerospace and Electronic Systems, 2019, 55(5):2543-2555
[7]Seokwon Lee, Jihoon Lee, Somang Lee, et al.Sliding mode guidance and control for UAV carrier landing[J].IEEE Transactions on Aerospace and Electronic Sys-tems, 2019, 55(2):951-966
[8]CHEN C, TAN W Q, QU X J, et al.A fuzzy human pilot model of longitudinal control for a carrier landing task[J].IEEE Transactions on Aerospace and Electronic Sys-tems, 2018, 54(1):453-466
[9]H.Sadat-Hoseini,SFazelzadeh,A. Rasti,et al. Final approach and flare control of a flexible aircraft in cross-wind landings[J].Journal of Guidance Control and Dy-namics, 2013, 36(4):946-957
[10]马国庆, 田秋扬, 朱波, 胡天江.孪生场景驱动的固定翼无人机进场自主着陆控制[J].控制理论与应用, 2024, 41(9):1664-1675
[11]MA G Q, TIAN Q Y ZHU B, HU T J.Twin-scenario driven autolanding control for fixed-wing UAVs[J].Control Theory & Applications, 2024, 41(9):1664-1675
[12]Jo?o C.Figueira,Sean Bazzocchi,Stephen Warwick and Afzal SulemanNonlinear aero-propulsive modeling for fixed-wing eVTOL UAV from flight test data[J].Jour-nal of Aircraft, 2025, 62(2):300-312
[13]CAO R, LU K L, LIU Y B.An interpretable data-driven learning approach for nonlinear aircraft systems with noisy interference[J].IEEE Transactions on Aerospace and Electronic Systems, 2025, 61(1):182-194
[14]王宏伦, 裴云峰, 倪少波.飞行器无动力应急着陆域和着陆轨迹设计[J].航空学报, 2014, 35(5):1404-1415
[15]WANG H L, PEI Y F, NI S B, et al.Design of emergen-cy landing region and landing trajectory for unpowered Aircraft[J].Acta Aeronautica et Astronautica Sinica, 2014, 35(5):1404-1415
[16]PAN Z L, PI Y J.Design of an automatic landing strate-gy for fixed-wing aircraft based on longitudinal and lateral collaborative hierarchical control[J].Aerospace Sci-ence and Technology, 2025, 163(163):110217-1-110217-15
[17]Youji Liguni, Hideo Akiyoshi, Norihiko Adachi, et al.An intelligent landing system based on human skill model[J].IEEE Transactions on Aerospace and Electronic Sys-tems, 1998, 34(3):877-881
[18]DUAN H B, CHEN L, ZENG Z G.Automatic landing for carrier-based aircraft under the conditions of deck motion and carrier airwake disturbances[J].IEEE Trans-actions on Aerospace and Electronic Systems, 2022, 58(6):5276-5291
[19]D.A. Bristow,MTharayil,A.G. Alleyne. A survey of iterative learning control[J].IEEE Control Systems Mag-azine, 2006, 26(3):96-114
[20]LI X F, HUANG D Q, CHU B, XU J X.Robust itera-tive learning control for systems with norm-bounded un-certainties[J].International Journal of Robust and Non-linear Control, 2015, 26(4):697-718
[21]Jeroen van de Wijdeven, Tijs Donkers, Okko Bosgra.Iterative learning control for uncertain systems: robust monotonic convergence analysis[J].Automatica, 2009, 45(45):2383-2391
[22]LI X F, XU J X, HUANG D Q.An iterative learning control approach for linear systems with randomly vary-ing trial lengths[J].IEEE Transactions on Automatic Control, 2014, 59(7):1954-1960
[23]XIA J K, ZHANG R Q, LI Y A, HUANG D Q, LI X F.Iterative learning control based on dynamic time warping for a class of discrete-time nonlinear systems with vary-ing trial lengths and terminus constraint[J].International Journal of Robust and Nonlinear Control, 2023, 33(33):3146-3163
[24]郭晓临, 刘洋, 林娜, 池荣虎.基于扩展状态观测器的量化无模型自适应迭代学习控制[J].控制理论与应用, 2025, 42(2):253-261
[25]GUO X L, LIU Y, LIN N, CHI R H.Extended state ob-server-based quantitative model-free adaptive iterative learning control[J].Control Theory & Applications, 2025, 42(2):253-261
[26]Shuli Lv, Pengda Mao, and Quan Quan.Mean-field based time-optimal spatial iterative learning within a vir-tual tube[J].IEEE Control Systems Letters, 2024, 8(8):2021-2026
[27]张凡, 孟德元, 惠宇.面向超机动飞行的迭代学习控制研究综述[J].空天技术, 2022, 6(6):2-23
[28]ZHANG F, MENG D Y, HUI Y.Review of iterative learning control for super-maneuver flight[J].Aerospace Technology, 2022, 6(6):2-23
[29]Randal W.Beard, Timothy W. McLain. Small unmanned aircraft: theory and practice [M]. New Jersey: Princeton University Press, 2012: 95-118.
[30]S.Baldi,D,SUN D P,Xia X,ZHOU G P,LIU D. Ar-duPilot-based adaptive autopilot: architecture and soft-ware-in-the-loop experiments[J].IEEE Transaction on Aerospace and Electronic Systems, 2022, 58(5):4473-4485
[31]Sun M X, LI W, WU L W.Dead-beat terminal sliding-mode control: a guaranteed attractiveness approach[J].Automatica, 2019, 129(129):109609-1-109609-9
[32]LV S L, Y.GAO Y, CHE J X and QUAN Q. Autono-mous Drone Racing: Time-Optimal Spatial Iterative Learning Control within a Virtual Tube [C]. IEEE Inter-national Conference on Robotics and Automation (ICRA), 2023, pp: 3197-3203.
文章导航

/