An AutoML based trajectory optimization method for long-distance spacecraft pursuit-evasion game

doi:10.23919/JSEE.2023.000060

Journal of Systems Engineering and Electronics ›› 2023, Vol. 34 ›› Issue (3): 754-765.doi: 10.23919/JSEE.2023.000060

• • 上一篇

收稿日期:2022-03-31 出版日期:2023-06-15 发布日期:2023-06-30

An AutoML based trajectory optimization method for long-distance spacecraft pursuit-evasion game

Fuyunxiang YANG(), Leping YANG(), Yanwei ZHU()

¹ College of Aerospace Science and Engineering, National University of Defense Technology, Changsha 410073, China

Received:2022-03-31 Online:2023-06-15 Published:2023-06-30
Contact: Yanwei ZHU E-mail:yangfuyunxiang@nudt.edu.cn;ylp_1964@163.com;zywnudt@163.com
About author:
YANG Fuyunxiang was born in 1996. He received his B.S. degree from National University of Defense Technology (NUDT), Changsha, China, in 2018. He is a Ph.D. candidate with the College of Aerospace Science and Engineering, NUDT. His research interests are aerospace dynamics and control, application of artificial intelligence to the control of astronautic systems. E-mail: yangfuyunxiang@nudt.edu.cn

YANG Leping was born in 1964. He received his B.S. and M.S. degrees from National University of Defense Technology (NUDT), Changsha, China, in 1984 and 1987, respectively. He is a professor with the College of Aerospace Science and Engineering, NUDT. His research interests include aerospace dynamics and control, aerospace mission planning and simulation. E-mail: ylp_1964@163.com

ZHU Yanwei was born in 1981. He received his B.S., M.S., and Ph.D. degrees from National University of Defense Technology (NUDT), Changsha, China, in 2002, 2004 and 2009 respectively. He is a professor with the College of Aerospace Science and Engineering, NUDT. His research interests include aerospace dynamics and control, aerospace mission planning and simulation. E-mail: zywnudt@163.com
Supported by:
This work was supported by the National Defense Science and Technology Innovation program (18-163-15-LZ-001-004-13)

摘要/Abstract

Abstract:

Current successes in artificial intelligence domain have revitalized interest in spacecraft pursuit-evasion game, which is an interception problem with a non-cooperative maneuvering target. The paper presents an automated machine learning (AutoML) based method to generate optimal trajectories in long-distance scenarios. Compared with conventional deep neural network (DNN) methods, the proposed method dramatically reduces the reliance on manual intervention and machine learning expertise. Firstly, based on differential game theory and costate normalization technique, the trajectory optimization problem is formulated under the assumption of continuous thrust. Secondly, the AutoML technique based on sequential model-based optimization (SMBO) framework is introduced to automate DNN design in deep learning process. If recommended DNN architecture exists, the tree-structured Parzen estimator (TPE) is used, otherwise the efficient neural architecture search (NAS) with network morphism is used. Thus, a novel trajectory optimization method with high computational efficiency is achieved. Finally, numerical results demonstrate the feasibility and efficiency of the proposed method.

Key words: pursuit-evasion, different game, trajectory optimization, automated machine learning (AutoML)

. [J]. Journal of Systems Engineering and Electronics, 2023, 34(3): 754-765.

Fuyunxiang YANG, Leping YANG, Yanwei ZHU. An AutoML based trajectory optimization method for long-distance spacecraft pursuit-evasion game[J]. Journal of Systems Engineering and Electronics, 2023, 34(3): 754-765.

图/表 15

参考文献 28

1	SHEN H X, CASALINO L Revisit of the three-dimensional orbital pursuit-evasion game. Journal of Guidance, Control and Dynamics, 2018, 41 (8): 1820- 1828.
2	LI Z Y, ZHU H, YANG Z, et al A dimension-reduction solution of free-time differential games for spacecraft pursuit-evasion. Acta Astronautica, 2019, 163, 201- 210. doi: 10.1016/j.actaastro.2019.01.011
3	LI Z Y, ZHU H, YANG Z, et al Saddle point of orbital pursuit-evasion game under J₂-perturbed dynamics . Journal of Guidance, Control and Dynamics, 2020, 43 (9): 1733- 1739. doi: 10.2514/1.G004459
4	ISAACS R. Differential games. New York: Wiley, 1965.
5	LUO Y Z, LI Z Y, ZHU H Survey on spacecraft orbital pursuit-evasion differential games. Scientia Sinica Technologica, 2020, 50 (12): 1533- 1545. doi: 10.1360/SST-2019-0174
6	CHENG L, WANG Z B, JIANG F H, et al Real-time optimal control for spacecraft orbit transfer via multiscale deep neural networks. IEEE Trans. on Aerospace and Electronic Systems, 2019, 55 (5): 2436- 2450. doi: 10.1109/TAES.2018.2889571
7	WU Q C, LI B, LI J, et al Solution of infinite time domain spacecraft pursuit strategy based on deep neural network. Aerospace Control, 2019, 37 (6): 13- 18,58.
8	GEORGE B C. Optimal and robust neural network controllers for proximal spacecraft maneuvers. Dayton: Air Force Institute of Technology, 2019.
9	ZENG X, ZHU Y W, YANG L P, et al A guidance method for coplanar orbital interception based on reinforcement learning. Journal of Systems Engineering and Electronics, 2021, 32 (4): 927- 938. doi: 10.23919/JSEE.2021.000079
10	YIN S S, LI J, CHENG L Low-thrust spacecraft trajectory optimization via a DNN-based method. Advances in Space Research, 2020, 66 (7): 1635- 1646. doi: 10.1016/j.asr.2020.05.046
11	YAO Q M, WANG M S, CHEN Y Q, et al. Taking human out of learning applications: a survey on automated machine learning. https://arxiv.org/abs/1810.13306v4.
12	FRANK H, LARS K, JOAQUIN V. Automated machine learning methods, systems, challenges. Berlin: Springer, 2019.
13	YU T, ZHU H. Hyper-parameter optimization: a review of algorithms and applications. https://arxiv.org/abs/2003.05689v1.
14	WISTUBA M, SCHILLING N, SCHMIDT-THIEME L Scalable Gaussian process-based transfer surrogates for hyperparameter optimization. Machine Learning, 2018, 107 (1): 43- 78. doi: 10.1007/s10994-017-5684-y
15	CHRIS T, FRANK H, HOLGER H, et al. Auto-WEKA: combined selection and hyperparameter optimization of classification algorithms. Proc. of the 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 2013: 847−855.
16	JAMES B, REMI B, YOSHUA B, et al. Algorithms for hyper-parameter optimization. Proc. of the 24th International Conference on Neural Information Processing Systems, 2011: 2546−2554.
17	KRIZHEVSKY A, SUTSKEVER I, HINTON G E Imagenet: classification with deep convolutional neural networks. Communications of the ACM, 2017, 60 (6): 84- 90. doi: 10.1145/3065386
18	HE X, ZHAO K Y, CHU X W AutoML: a survey of the state-of-the-art. Knowledge-Based Systems, 2021, 212, 106622. doi: 10.1016/j.knosys.2020.106622
19	STORK J, ZAEFFERER M, BARTZ-BEIELSTEIN T. Improving neuroevolution efficiency by surrogate model-based optimization with phenotypic distance kernels. https://arxiv.org/abs/1902.03419v1.
20	CAMERO A, WANG H, ALBA E, et al. Bayesian neural architecture search using a training-free performance metric. https://arxiv.org/abs/2001.10726.
21	JIN H F, SONG Q Q, HU X. Auto-Keras: an efficient neural architecture search system. Proc. of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2019: 1946−1956.
22	BASAR T, OLSDER G. Dynamic non-cooperative game theory. Philadelphia: Society for Industrial and Applied Mathematics, 1999.
23	JIANG F H, BAOYIN H X, LI J F Practical techniques for low-thrust trajectory optimization with homotopic approach. Journal of Guidance, Control, and Dynamics, 2012, 35 (1): 245- 258. doi: 10.2514/1.52476
24	FAMILI A, SHEN W M, WEBER R, et al Data preprocessing and intelligent data analysis. Intelligent Data Analysis, 1997, 1 (1): 3- 23. doi: 10.3233/IDA-1997-1102
25	FRANK H, HOLGER H H, KEVIN L B. Sequential model-based optimization for general algorithm configuration. Proc. of the 5th International Conference on Learning and Intelligent Optimization, 2011: 507−523.
26	DONALD R J A taxonomy of global optimization methods based on response surfaces. Journal of Global Optimization, 2001, 21 (4): 345- 383. doi: 10.1023/A:1012771025575
27	ZENG Z P, ANTHONY K H T, WANG J Y, et al Comparing stars: on approximating graph edit distance. Proceedings of the VLDB Endowment, 2009, 2 (1): 25- 36. doi: 10.14778/1687627.1687631
28	PETER A, NICOLO C B, PAUL F Finite-time analysis of the multiarmed bandit problem. Machine Learning, 2002, 47 (2): 235- 256.

Element	min	max
Semi-major axis	6.47	6.78
Eccentricity	0	0.3
Inclination/rad	0	π/36
Right ascension of the ascending node/rad	0	2π
Argument of perigee/rad	0	2π
Ture anomaly/rad	0	2π

Data name	Resnet18	HPO net	NAS net
Total parameter	2 857 260	888 588	545 103
Trainable parameter	2 852 332	885 708	545 100
Non-trainable parameter	4 928	2 880	3
Learning rate	0.0010	0.0059	0.0024
Dropout rate	0.3000	0.4958	0.2178
Batch size	64	128	64
Max epoch	600	600	600
Loss function	MSE	MSE	MSE

Parameter	SADE	Resnet18	HPO	NAS
Average time consumption/s	183.9439	0.0038	0.0031	0.0016
Average computation error	0.0083	0.0082	0.0144	0.0072
Standard deviation of error	0.0072	0.0162	0.0232	0.0172

Methods	Final time	Error	Computation time/s
SADE	0.2434	0.0045	183.3176
Resnet18	0.2571	0.0037	0.0037
HPO	0.2520	0.0057	0.0031
NAS	0.2138	0.0002	0.0016

An AutoML based trajectory optimization method for long-distance spacecraft pursuit-evasion game

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 15

参考文献 28

相关文章 0

编辑推荐

Metrics

本文评价