School of Aeronautics and Astronautics,Zhejiang University,Hangzhou 310027,China
Show less
History+
Received
Published Online
2026-05-25
2026-11-03
PDF
(989KB)
Abstract
To address the significant rigid-flexible coupling effects and the difficulty of real-time planning under limited on-board computational resources during large-angle attitude maneuvers of flexible spacecraft, an active vibration suppression attitude planning method based on a dual-branch temporal neural planner was proposed. A controller-aware strategy to offline solve optimal trajectories was introduced, generating a dynamics-constrained expert dataset, based on this, a dual-branch temporal neural planner was designed, where the TCN branch extracts temporal vibration features and the Multilayer Perceptron(MLP)branch resolves instantaneous maneuver targets, jointly trained via a composite loss function. Once trained, the planner takes historical state sequences and maneuver targets as input and outputs attitude planning commands in real time. Simulation tests demonstrate that the proposed method effectively suppresses flexible vibrations, significantly shortens the settling time, and substantially reduces residual vibrations and control energy consumption.
Guitao Yao, Fanyu Zhao.
Active Vibration Suppression Attitude Planning for Flexible Spacecraft Based on a Dual-Branch Temporal Neural Planner.
Journal of Deep Space Exploration, 2026, 13 (4) : 323-334 DOI:10.3724/j.issn.2096-9287.2026.20260041
求解最优姿态规划问题的主要手段包括间接法与直接法[10]。间接法基于庞特里亚金极大值原理,求解精度高但收敛区间窄,对协态变量初始猜测极其敏感;直接法(如伪谱法、直接转录法)通过离散化状态与控制变量,将问题转化为非线性规划问题[11],然后使用非线性求解器如 IPOPT[12](Interior Point OPTimizer)进行求解,具备较强的约束处理能力。然而,随着航天器模型复杂度的增加,直接法要消耗大量的计算资源[13]。在星载处理器资源受限的情况下,这类算法难以满足高动态任务的实时重规划需求。
Meirovitch & KwakM K. Control of flexible spacecraft with time-varying configuration. Journal of Guidance, Control, and Dynamics15(2), 314–324 (1992).
[2]
JinE & SunZ W. Design of nonlinear PID and PI attitude controller for a flexible spacecraft. Control Theory Application26(5), 540–545 (2009).
[3]
JinY QLiuX & HouC Z. Sliding-mode attitude tracking control for a flexible spacecraft with parametric uncertainty. Control Theory and Applications26(3), 299–304 (2009).
[4]
AzimiMShahravi & FardK M. Modeling and vibration suppression of flexible spacecraft using higher-order sandwich panel theory. International Journal of Acoustics and Vibration22(2), 143–151 (2017).
[5]
LiALiu & ShiY. Adaptive sliding mode attitude tracking control for flexible spacecraft systems based on the Takagi-Sugeno fuzzy modelling method. Acta Astronautica175, 570–581 (2020).
[6]
WuC & ChenB S. Attitude control of spacecraft: mixed H/H approach. Journal of Guidance, Control, and Dynamics24(4), 755–766 (2001).
[7]
YuY NMeng & LiK. Robust control of flexible spacecraft during large-angle attitude maneuver. Journal of Guidance, Control, and Dynamics37(3), 1027–1033 (2014).
[8]
FanL SHuang & SunL. Robust attitude control for a rigid-flexible-rigid microsatellite with multiple uncertainties and input saturations. Aerospace Science and Technology95, 1–11 (2019).
[9]
SinghoseWEloundou & LawrenceJ. Command generation for flexible systems by input shaping and command smoothing. Journal of Guidance, Control, and Dynamics33(6), 1697–1707 (2010).
[10]
BettsJ T Survey of numerical methods for trajectory optimization. Journal of Guidance, Control, and Dynamics21(2), 193–207 (1998).
[11]
Betts J T. Practical methods for optimal control and estimation using nonlinear programming[M]. Philadelphia: Society for Industrial and Applied Mathematics, 2010.
[12]
BieglerL & ZavalaV M. Large-scale nonlinear programming using IPOPT: An integrating framework for enterprise-wide dynamic optimization. Computers & Chemical Engineering33(3), 575–582 (2009).
[13]
EnrightP & ConwayB A. Discrete approximations to optimal trajectories using direct transcription and nonlinear programming. Journal of Guidance, Control, and Dynamics15(4), 994–1002 (1992).
[14]
Sánchez-sánchez C, Izzo D, Hennes D. Learning the optimal state-feedback using deep networks[C]//Proceedings of 2016 IEEE Symposium Series on Computational Intelligence(SSCI). Athens, Greece: IEEE, 2016.
[15]
Sánchez-sánchez & IzzoD. Real-time optimal control via deep neural networks: Study on landing problems. Journal of Guidance, Control, and Dynamics41(5), 1122–1135 (2018).
[16]
BiggsJ & FournierH. Neural-network-based optimal attitude control using four impulsive thrusters. Journal of Guidance, Control, and Dynamics43(2), 299–309 (2020).
[17]
ChengLWangZ & SongY. Real-time optimal control for irregular asteroid landings using deep neural networks. Acta Astronautica170, 66–79 (2020).
[18]
Shi & WangZ B. Onboard generation of optimal trajectories for hypersonic vehicles using deep learning. Journal of Spacecraft and Rockets58(2), 400–414 (2021).
[19]
You S X, Wan C H, Dai R, et al. Learning-based optimal control for planetary entry, powered descent and landing guidance[C]//Proceedings of AIAA SciTech 2020 forum. Orlando, FL: AIAA, 2020.
[20]
Bai S J, Kolter J Z, Koltun V. An empirical evaluation of generic convolutional and recurrent networks for sequence modeling[EB/OL].(2018-4-19)[2026-05-17]. https: //arxiv.org/abs/1803.01271.
[21]
Junkins J L, Kim Y. Introduction to dynamics and control of flexible structures[M]. Reston: American Institute of Aeronautics and Astronautics, 2000.
[22]
XiaoBYin & KaynakO. Attitude stabilization control of flexible satellites with high accuracy: an estimator-based approach. IEEE/ASME Transactions on Mechatronics22(1), 349–358 (2017).
[23]
Hu & MaG. Variable structure control and active vibration suppression of flexible spacecraft during attitude maneuver. Aerospace Science and Technology9(4), 307–317 (2005).
[24]
ChowSChang & HollingerG A. Parallelized control-aware motion planning with learned controller proxies. IEEE Robotics and Automation Letters8(4), 2237–2244 (2023).
[25]
TognonMCataldi & ChavezH A T. Control-aware motion planning for task-constrained aerial manipulation. IEEE Robotics and Automation Letters3(3), 2478–2484 (2018).
[26]
Markley F L, Crassidis J L. Fundamentals of Spacecraft Attitude Determination and Control[M]. 1 ed. New York: Springer, 2014.
[27]
Wächter & BieglerL T. On the implementation of an interior-point filter line-search algorithm for large-scale nonlinear programming. Mathematical Programming106(1), 25–57 (2006).
Dietterich T G. Machine learning for sequential data: a review[C]//Proceedings of Joint IAPR International Workshops SSPR 2002 and SPR 2002. Windsor, Ontario, Canada: Springer, 2002.
[30]
KaelblingL PLittmanM & CASSANDRAA R. Planning and acting in partially observable stochastic domains. Artificial Intelligence101(1), 99–134 (1998).
[31]
Takens F. Detecting strange attractors in turbulence[C]//Proceedings of Dynamical Systems and Turbulence, Warwick 1980. Berlin, Heidelberg: Springer, 1981.
[32]
SauerTYorkeJ & CasdagliM. Embedology. Journal of Statistical Physics65(3), 579–616 (1991).
[33]
DeyleE & SugiharaG. Generalized theorems for nonlinear state space reconstruction. PLoS ONE6(3), 1–8 (2011).
[34]
LuschBKutzJ & BruntonS L. Deep learning for universal linear embeddings of nonlinear dynamics. Nature Communications9(1), 1–10 (2018).
[35]
Bakarji J, Champion K M, Nathan K J, et al. Discovering governing equations from partial measurements with deep delay autoencoders[J/OL]. Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences, 2023, 479(2276): 20230422. https: //royalsocietypublishing.org/rspa/article/479/2276/20230422/82633/Discovering-governing-equations-from-partial.
[36]
Pascanu R, Mikolov T, Bengio Y. On the difficulty of training recurrent neural networks[C]// Proceedings of the 30th International Conference on Machine Learning. Atlanta, Georgia, USA: PMLR, 2013.
[37]
Calderon J, Berman G J. Inferring the time-varying coupling of dynamical systems with temporal convolutional autoencoders[J/OL]. eLife, 2024, 13: RP100692. https: //elifesciences.org/reviewed-preprints/100692v1#tab-content.
[38]
ShannonC E Communication in the presence of noise. Proceedings of the IRE37(1), 10–21 (1949).
[39]
Oppenheim A V, Schafer R W. Discrete-time signal processing[M]. 3rd Edition, USA: Prentice Hall, 2010.
[40]
HuynhD Q Metrics for 3D rotations: comparison and analysis. Journal of Mathematical Imaging and Vision35(2), 155–164 (2009).
[41]
AnderssonJGillis & HornG. CasADi—a software framework for nonlinear optimization and optimal control. Mathematical Programming Computation11(1), 1–36 (2018).
[42]
PaszkeAGross & MassaF. Pytorch: An imperative style, high-performance deep learning library. Advances in Neural Information Processing Systems32, 8026–8037 (2019).
Rights & permissions
The Author(s) 2026. Published by Higher Education Press. This is an Open Access article distributed under the terms of the CC BY license (https://creativecommons.org/licenses/by/4.0/).