基于强化学习的无角速度多无人机固定时间姿态协同控制

doi:10.12305/j.issn.1001-506X.2026.11.29

摘要/Abstract

摘要：

针对角速度传感器不可用的多无人机协同控制问题，提出一种基于强化学习的固定时间最优姿态控制方法。首先，设计基于神经网络的固定时间观测器，仅需姿态测量信息即可精确重构角速度信号，解决了传感器缺失状态估计难题。其次，结合强化学习理论，提出一种基于分布式观测器的姿态一致性最优控制器，实现跟踪误差的快速收敛。进一步地，基于Lyapunov稳定性理论，严格证明了所提控制方案的固定时间收敛性。最后，仿真结果表明，所提方法可在固定时间内精准重构角速度信号，稳定实现多无人机角度跟踪与姿态一致性控制，降低了传感器缺失对多无人机姿态协同控制的不利影响，同时验证了所设计控制器的有效性和鲁棒性，进一步说明了其在实际应用中的可行性。

关键词: 强化学习, 同步控制, 固定时间控制, 固定时间神经网络观测器

Abstract:

Aiming at the problem of cooperative control of multiple quadrotors when the angular velocity sensors are unavailable, a fixed-time optimal attitude control method based on the reinforcement learning is proposed. Firstly, a fixed-time observer based on a neural network is designed. It can accurately reconstruct the angular velocity signal only with the attitude measurement information, thus solving the difficult problem of state estimation in the absence of sensors. Secondly, combining the reinforcement learning theory, an attitude consensus optimal controller based on a distributed observer is proposed, which achieves the rapid convergence of the tracking error. Furthermore, based on the Lyapunov stability theory, the fixed-time convergence of the proposed control scheme is rigorously proven. Finally, the simulation results demonstrate that the proposed method can accurately reconstruct the angular velocity signals within a fixed time, stably achieve multi-quadrotor angle tracking and attitude consensus control, and mitigate the adverse impacts of sensor absence on the attitude cooperative control of multi-quadrotor systems. Meanwhile, the effectiveness and robustness of the designed controller are verified, which further illustrates its feasibility in practical applications.

Key words: reinforcement learning, synchronization control, fixed-time control, fixed-time neural network observer

中图分类号:

TP 273.1

杨一帆, 王利楠, 温广辉. 基于强化学习的无角速度多无人机固定时间姿态协同控制[J]. 系统工程与电子技术, 2026, 48(11): 3999-4007.

Yifan Yang, Linan Wang, Guanghui Wen. Fixed-time attitude cooperative control for quadrotors without angular velocity measurement based on reinforcement learning[J]. Systems Engineering and Electronics, 2026, 48(11): 3999-4007.

图/表 9

图1

表1

表2

表3

表4

图2

图3

图4

图5

参考文献 31

1	Li Y, Li Z, Yu X H, et al. Attitude tracking control for quadrotors under unknown disturbances based on dual-loop nonlinear command filters[J]. IEEE Trans. on Circuits and Systems II: Express Briefs, 2024, 71 (5): 2769. doi: 10.1109/tcsii.2024.3351686
2	Zhang J, Li F S, Xin Y L, et al. Trajectory generation for quadrotors air parallel docking with time-varying avoidance constraints[J]. IEEE Robotics and Automation Letters, 2025, 10 (12): 12820. doi: 10.1109/LRA.2025.3625489
3	Zhu C J, Chen J C, Iwasaki M, et al. Event-triggered deep learning control of quadrotors for trajectory tracking[J]. IEEE Trans. on Industrial Electronics, 2023, 71 (3): 2726.
4	Zhao W B, Liu H, Lü J H, et al. Data-driven optimal formation control for multiple nonlinear quadrotors with switching topologies[J]. IEEE Trans. on Vehicular Technology, 2023, 74 (2): 2538.
5	Yang Y H, Mei J, Deng C, et al. Distributed event-triggered formation control of mass with parametric uncertainties under a directed graph: application to multiple quadrotors[J]. IEEE Trans. on Industrial Electronics, 2024, 72 (7): 7343.
6	Lapandi D, Xie F Z, Verginis C K, et al. Meta-learning augmented MPC for disturbance-aware motion planning and control of quadrotors[J]. IEEE Control Systems Letters, 2024, 8, 3045. doi: 10.1109/LCSYS.2024.3520023
7	Asignacion A, Suzuki S. Frequency-based robust control for micro quadrotors flying in narrow vertical spaces[J]. IEEE Access, 2024, 12, 141742. doi: 10.1109/ACCESS.2024.3464492
8	王利楠, 温广辉, 伊枭剑. 状态约束下的多飞行器有限时间姿态一致性控制[J]. 现代防御技术, 2024, 52 (2): 124. doi: 10.3969/j.issn.1009-086x.2024.02.014
9	耿泽, 黄炎焱, 张寒. 基于火炮转移路径预测的无人机集群反炮兵搜索路径规划[J]. 系统工程与电子技术, 2025, 47 (4): 1222. doi: 10.12305/j.issn.1001-506X.2025.04.19
10	Wen G H, Wang L N, Zhao D, et al. Attitude estimation for rigid aircraft with time-varying gyro bias: a finite-time complementary filtering approach[J]. Guidance, Navigation and Control, 2025, 5 (4): 445. doi: 10.1142/s2737480725500268
11	桂洋, 郑柏超, 高鹏. 基于NESO-LFDC的四旋翼无人机滑模姿态控制[J]. 系统工程与电子技术, 2024, 46 (3): 1075. doi: 10.12305/j.issn.1001-506X.2024.03.34
12	付俊杰, 林潇坤, 温广辉. 基于高阶控制障碍函数的多固定翼无人机鲁棒避障安全编队跟踪控制[J]. 机器人, 2025, 47 (1): 85.
13	丁岩, 于志刚. 考虑输入输出受限的无人机自适应滑模容错控制[J]. 系统工程与电子技术, 2020, 42 (10): 2340. doi: 10.3969/j.issn.1001-506X.2020.10.24
14	Cui B, Xia Y Q, Liu K, et al. Velocity-observer-based distributed finite-time attitude tracking control for multiple uncertain rigid spacecraft[J]. IEEE Trans. on Industrial Informatics, 2019, 16 (4): 2509.
15	Xiao B, Zhang H C, Chen Z Y, et al. Fixed-time fault-tolerant optimal attitude control of spacecraft with performance constraint via reinforcement learning[J]. IEEE Trans. on Aerospace and Electronic Systems, 2023, 59 (6): 7715.
16	Yang H Y, Hu Q L, Dong H Y, et al. ADP-based spacecraft attitude control under actuator misalignment and pointing constraints[J]. IEEE Trans. on Industrial Electronics, 2021, 69 (9): 9342.
17	Xue S, Luo B, Liu D R, et al. Event-triggered ADP for tracking control of partially unknown constrained uncertain systems[J]. IEEE Trans. on Cybernetics, 2021, 52 (9): 9001.
18	Hu J P, Yang G, Hou Z C, et al. A nearer optimal and faster trained value iteration ADP for discrete-time nonlinear systems[J]. IEEE Access, 2021, 9, 14933.
19	Yang H Y, Hu Q L, Dong H Y, et al. Optimized data-driven prescribed performance attitude control for actuator saturated spacecraft[J]. IEEE/ASME Trans. on Mechatronics, 2023, 28 (3): 1616.
20	Wang M L, Wu B L, Geng Y H, et al. Distributed approximate optimal attitude control of combined spacecraft with parameters identification[J]. IEEE Trans. on Aerospace and Electronic Systems, 2023, 60 (2): 1784.
21	Zhao W B, Liu H, Lewis F L. Data-driven fault-tolerant control for attitude synchronization of nonlinear quadrotors[J]. IEEE Trans. on Automatic Control, 2021, 66 (11): 5584.
22	Zhao J Y, Li F F. Attitude control for a rigid-body with dynamic disturbance based on angular velocity observer[J]. IEEE Access, 2022, 10, 36877.
23	Zou A M, Kumar K D, Hou Z G. Attitude coordination control for a group of spacecraft without velocity measurements[J]. IEEE Trans. on Control Systems Technology, 2011, 20 (5): 1160.
24	Zhu W W, Yang Y K, Tian B L, et al. Velocity-free attitude control for flexible spacecraft with global finite-time convergence[J]. IEEE Trans. on Circuits and Systems II: Express Briefs, 2024, 71 (9): 4226.
25	Ye D, Zou A M, Sun S X, et al. A predefined-time extended-state observer- based approach for velocity-free attitude control of spacecraft[J]. IEEE Trans. on Aerospace and Electronic Systems, 2023, 59 (6): 8051.
26	Yu Z L, Li Y H, Lv M L, et al. Fixed-time adaptive fuzzy fault-tolerant attitude control for tailless aircraft without angular velocity measurements[J]. IEEE Trans. on Fuzzy Systems, 2024, 32 (4): 2458.
27	Ning K, Wu B L, Wang D W. Velocity observer-based event-triggered adaptive fuzzy attitude takeover control of spacecraft with quantized quaternion[J]. IEEE Trans. on Aerospace and Electronic Systems, 2023, 59 (4): 4168.
28	Yang S Y, Li B, Liu H, et al. Adaptive dynamic programming-based fault tolerant control for a quadrotor UAV attitude stabilization[C]//35th Chinese Control and Decision Conference, 2023: 2612.
29	Ba D S, Li Y X, Tong S C. Fixed-time adaptive neural tracking control for a class of uncertain nonstrict nonlinear systems[J]. Neurocomputing, 2019, 363, 273.
30	Bhasin S, Kamalapurkar R, JOHNSON M, et al. A novel actor-critic-identifier architecture for approximate optimal control of uncertain nonlinear systems[J]. Automatica, 2013, 49 (1): 82.
31	Al-tamimi A, Lewis F L, Abu-khalaf M. Discrete-time nonlinear HJB solution using approximate dynamic programming: convergence proof[J]. IEEE Trans. on Systems, Man, and Cybernetics, Part B （Cybernetics）, 2008, 38 (4): 943.

编号	$ {\varpi }_{\phi } $/（°）	$ {\varpi }_{\theta } $/（°）	$ {\varpi }_{\psi } $/（°）
1	17.2	11.5	5.7
2	51.6	34.4	17.2
3	−85.9	−57.3	−28.6
4	137.5	91.7	45.8

编号	$ \phi $/（°）	$ \theta $/（°）	$ \psi $/（°）
1	5.7	11.4	−2.8
2	6.0	12.0	−2.9
3	5.4	10.8	−2.7
4	6.8	13.7	−3.4

编号	$ \hat{\phi } $/（°）	$ \hat{\theta } $/（°）	$ \hat{\psi } $/（°）
1	5.7	11.4	−2.8
2	5.7	11.4	−2.8
3	5.7	11.4	−2.8
4	5.7	11.4	−2.8

编号	$ \widehat{\dot{\phi } } $/（°/s）	$ \widehat{\dot{\theta } } $/（°/s）	$ \widehat{\dot{\psi } } $/（°/s）
1	0	57.2	114.6
2	114.6	171.8	286.4
3	−57.2	57.2	57.2
4	0	0	0

[1]	赵立阳, 吴金平, 杨静, 张瑞. 基于深度强化学习的多智能体集中式作战行动指挥决策方法[J]. 系统工程与电子技术, 2026, 48(9): 3089-3103.
[2]	张杰, 朱峰, 王超, 李东, 陈志群, 吕天启. 逻辑链构建下基于深度强化学习的多智能体协同算法优化[J]. 系统工程与电子技术, 2026, 48(9): 3126-3135.
[3]	张倬境, 盛永智, 龚孝龙. 深度强化学习驱动的多约束分数阶滑模制导律[J]. 系统工程与电子技术, 2026, 48(9): 3167-3178.
[4]	刘锡国, 刘敏, 毛忠阳, 徐盟飞, 马鹏. 基于信息调节LSTM的抗干扰决策算法[J]. 系统工程与电子技术, 2026, 48(9): 3266-3273.
[5]	张杰, 王超, 刘玉, 李东, 陈志群, 吕天启. 对抗场景下基于多智能体联盟更新的指控逻辑链优化问题研究[J]. 系统工程与电子技术, 2026, 48(8): 2782-2789.
[6]	尹天然, 罗贺, 石悦, 强小蝶, 王国强. 基于强化混合遗传算法的卫星与无人机联合任务规划方法[J]. 系统工程与电子技术, 2026, 48(7): 2293-2306.
[7]	杨京礼, 王莹, 高天宇, 方晓彤. 面向故障传播风险抑制的舰船电子信息系统动态重构方法[J]. 系统工程与电子技术, 2026, 48(7): 2307-2318.
[8]	李强, 周荻, 李思远, 林玉荣. 基于元学习与预测控制的导弹自适应制导[J]. 系统工程与电子技术, 2026, 48(7): 2424-2433.
[9]	赵友雪, 周军, 冯振欣. 敏捷卫星固定时间三轴同步姿态控制器设计[J]. 系统工程与电子技术, 2026, 48(7): 2448-2456.
[10]	马赞, 刘禹彬, 白杰, 陈勇, 孙淑光. 基于分层STPA-MC的无人机智能避让适航安全风险评估[J]. 系统工程与电子技术, 2026, 48(6): 2000-2013.
[11]	贾振帅, 肖冰, 钱寒雨, 张哲宇. 基于多策略学习的航天器在轨观测机动决策方法[J]. 系统工程与电子技术, 2026, 48(5): 1590-1598.
[12]	张华卿, 张晓飞, 郝明瑞, 姜吉祥, 李闪. 鲁棒多智能体协同对抗策略离线强化学习[J]. 系统工程与电子技术, 2026, 48(5): 1670-1681.
[13]	张少辉, 李璐璐, 李亚飞, 吴庆顺, 李冠峰, 徐明亮. 舰载机弹药保障作业规划方法研究综述[J]. 系统工程与电子技术, 2026, 48(4): 1303-1321.
[14]	金志刚, 刘泽培, 武晓栋. 物联网安全与隐私保护中博弈论的应用综述[J]. 系统工程与电子技术, 2026, 48(3): 1072-1082.
[15]	杨鹏程, 杨清清, 高盈盈, 杨志伟, 杨克巍, 艾波. 基于强化学习的海上移动目标搜索路径规划[J]. 系统工程与电子技术, 2026, 48(2): 515-523.