基于指针网络架构的多星协同成像任务规划方法

doi:10.12305/j.issn.1001-506X.2025.07.18

摘要/Abstract

摘要：

随着卫星资源数量增加, 用户成像需求也在急剧扩大, 亟需加强多星协同成像任务规划研究, 提升卫星服务能力。本文基于深度强化学习对多星协同成像任务规划问题开展研究。首先, 在满足任务需求、卫星能力、时空约束基础上, 建立多星协同成像任务规划数学模型。然后, 设计一种基于指针网络的卫星任务规划算法, 利用指针网络机制对输入序列进行优化选择, 并通过Mask向量表征各类约束。最后, 仿真结果表明算法获得的平均任务收益比传统启发式算法和指针网络模型至少提高1.71%, 对于不同任务规模实例训练完成的算法, 其平均任务收益差最大不超过0.28%, 证明了算法的有效性和适用性。

关键词: 多星协同成像, 任务规划, 深度强化学习, 指针网络

Abstract:

With the increase in the number of satellite resources, user imaging demands are also rapidly expanding. There is an urgent need to strengthen research on multi-satellite coordinated imaging task planning to enhance satellite service capabilities. This paper conducts research on multi-satellite coordinated imaging task planning based on deep reinforcement learning. Firstly, a mathematical model for multi-satellite cooperative imaging task planning is established, taking into account task requirements, satellite capabilities, and spatiotemporal constraints. Then, a satellite task planning algorithm based on pointer network is designed. This algorithm employs the pointer network mechanism to optimize the selection of input sequences and utilizes a Mask vector to represent various constraint conditions. Finally, simulation results show that the algorithm achieves an average task benefit improvement of at least 1.71% compared to traditional heuristic algorithms and pointer network model. The average task benefit difference for algorithms trained on instances of different task scales is no more than 0.28%, demonstrating the effectiveness and applicability of the algorithm.

Key words: multi-satellite cooperative imaging, task planning, deep reinforcement learning, pointer network

中图分类号:

朱运豆, 孙海权, 胡笑旋. 基于指针网络架构的多星协同成像任务规划方法[J]. 系统工程与电子技术, 2025, 47(7): 2246-2255.

Yundou ZHU, Haiquan SUN, Xiaoxuan HU. Multi-satellite cooperative imaging task planning method based on pointer network architecture[J]. Systems Engineering and Electronics, 2025, 47(7): 2246-2255.

图/表 11

图1

图2

图3

图4

表1

表2

图5

图6

图7

表3

图8

参考文献 31

1	GUO H D , MICHAEL F G , ALESSANDRO A . Remote sensing satellites for digital earth[M]. Manual of Digital Earth, Singapore: springer, 2020.
2	李阳阳, 罗俊仁, 张万鹏, 等. 多星协同观测遗传-演进双层任务规划算法[J]. 系统工程与电子技术, 2024, 46 (6): 2044- 2053. doi: 10.12305/j.issn.1001-506X.2024.06.22
	LI Y Y , LUO J R , ZHANG W P , et al. Genetic-evolutionary bi-level mission planning algorithm for multi-satellite cooperative observation[J]. Systems Engineering and Electronics, 2024, 46 (6): 2044- 2053. doi: 10.12305/j.issn.1001-506X.2024.06.22
3	XU Y G , LIU X L , HE R J , et al. Multi-satellite scheduling framework and algorithm for very large area observation[J]. Acta Astronautica, 2020, 167, 93- 107.
4	LONG J , WU S M , HAN X D , et al. Autonomous task planning method for multi-satellite system based on a hybrid genetic algorithm[J]. Aerospace, 2023, 10 (1): 70.
5	PENG G S , SONG G P , XING L N , et al. An exact algorithm for agile earth observation satellite scheduling with time-dependent profits[J]. Computers & Operations Research, 2020, 120, 104946.
6	WANG J J , DEMEULEMEESTER E , QIU D S . A pure proactive scheduling algorithm for multiple earth observation satellites under uncertainties of clouds[J]. Computers & Operations Research, 2016, 74, 1- 13.
7	HU X X , ZHU W M , AN B , et al. A branch and price algorithm for EOS constellation imaging and downloading integrated scheduling problem[J]. Computers & Operations Research, 2019, 104, 74- 89.
8	KANDEPI R , SAINI H , GEORGE R K , et al. Agile earth observation satellite constellations scheduling for large area target imaging using heuristic search[J]. Acta Astronautica, 2024, 219, 670- 677.
9	周美玉, 印小冬, 刘聪, 等. 多星任务规划模型及算法[J]. 指挥信息系统与技术, 2023, 14 (3): 57- 64.
	ZHOU M Y , YIN X D , LIU C , et al. Multi-satellite task scheduling model and algorithm[J]. Command Information System and Technology, 2023, 14 (3): 57- 64.
10	ZHIBO E , SHI R H , GAN L , et al. Multi-satellites imaging scheduling using individual reconfiguration based integer coding genetic algorithm[J]. Acta Astronautica, 2021, 178, 645- 657.
11	HAN C , GU Y , WU G H , et al. Simulated annealing-based heuristic for multiple agile satellites scheduling under cloud coverage uncertainty[J]. IEEE Trans.on Systems, Man, and Cybernetics: Systems, 2022, 53 (5): 2863- 2874.
12	CHEN X Y , REINELT G , DAI G M , et al. Priority-based and conflict-avoidance heuristics for multi-satellite scheduling[J]. Applied Soft Computing, 2018, 69, 177- 191.
13	ZHAO X X , WANG Z K , ZHENG G . Two-phase neural combinatorial optimization with reinforcement learning for agile satellite scheduling[J]. Journal of Aerospace Information Systems, 2020, 17 (7): 346- 357.
14	WANG H N , LIU N , ZHANG Y Y , et al. Deep reinforcement learning: a survey[J]. Frontiers of Information Technology & Electronic Engineering, 2020, 21 (12): 1726- 1744.
15	BELLO I, PHAM H, LE Q V, et al. Neural combinatorial optimization with reinforcement learning[EB/OL]. [2024-06-11]. https://doiorg/10.48550/arXiv.1611. 09940.
16	NAZARI M, OROOJLOOY A, SNYDER L, et al. Reinforcement learning for solving the vehicle routing problem[C]//Proc. of the 32nd International Conference on Information Processing Systems, 2018.
17	WANG X W , WU G H , XING L N , et al. Agile earth observation satellite scheduling over 20 years: formulations, methods, and future directions[J]. IEEE Systems Journal, 2020, 15 (3): 3881- 3892.
18	LIU S K , YANG J . A satellite task planning algorithm based on a symmetric recurrent neural network[J]. Symmetry, 2019, 11 (11): 1373.
19	彭双, 伍江江, 陈浩, 等. 基于注意力神经网络的对地观测卫星星上自主任务规划方法[J]. 计算机科学, 2022, 49 (7): 242- 247.
	PENG S , WU J J , CHEN H , et al. Satellite onboard observation task planning based on attention neural network[J]. Computer Science, 2022, 49 (7): 242- 247.
20	CHEN J W , CHEN M , WEN J , et al. A heuristic construction neural network method for the time-dependent agile earth observation satellite scheduling problem[J]. Mathematics, 2022, 10 (19): 3498.
21	WU J , SONG B , ZHANG G T , et al. A data-driven improved genetic algorithm for agile earth observation satellite scheduling with time-dependent transition time[J]. Computers & Industrial Engineering, 2022, 174, 108823.
22	WANG X , WU J , SHI Z , et al. Deep reinforcement learning-based autonomous mission planning method for high and low orbit multiple agile Earth observing satellites[J]. Advances in Space Research, 2022, 70 (11): 3478- 3493.
23	LI P Y , WANG H Q , ZHANG Y X , et al. Mission planning for distributed multiple agile Earth observing satellites by attention-based deep reinforcement learning method[J]. Advances in Space Research, 2024, 74 (5): 2388- 2404.
24	WEI L N , CHEN Y N , CHEN M , et al. Deep reinforcement learning and parameter transfer based approach for the multi-objective agile earth observation satellite scheduling problem[J]. Applied Soft Computing, 2021, 110, 107607.
25	马一凡, 赵凡宇, 王鑫, 等. 基于改进指针网络的卫星对地观测任务规划方法[J]. 浙江大学学报(工学版), 2021, 55 (2): 395- 401.
	MA Y F , ZHAO F Y , WANG X , et al. Satellite earth observation task planning method based on improved pointer networks[J]. Journal of ZheJiang University (Engineering Science), 2021, 55 (2): 395- 401.
26	PETERS J , SCHAAL S . Natural Actor-Critic[J]. Neurocomputing, 2008, 71 (7/9): 1180- 1190.
27	KINGMA D P, BA J. Adam: a method for stochastic optimization[EB/OL]. [2024-06-11]. https://doi.org/10.48550/arXiv.2423.13370.
28	ZHANG J W , XING L N . An improved genetic algorithm for the integrated satellite imaging and data transmission scheduling problem[J]. Computers & Operations Research, 2022, 139, 105626.
29	丁祎男, 刘羽白, 王淑一, 等. 一种多目标变邻域模拟退火算法及成像星座任务规划方法[J]. 宇航学报, 2022, 43 (12): 1686- 1695.
	DING Y N , LIU Y B , WANG S Y , et al. A multi objective variable neighborhood simulated annealing algorithm and imaging constellation task planning method[J]. Journal of Astronautics, 2022, 43 (12): 1686- 1695.
30	ZHAO Y B , DU B , LI S . Agile satellite mission planning via task clustering and double-layer tabu algorithm[J]. Computer Modeling in Engineering & Sciences, 2020, 122 (1): 235- 257.
31	VINYALS O, FORTUNATO M, JAITLY N. Pointer networks[C]//Proc. of the 29th International Conference on Neural Information Processing Systems, 2015.

实例	经度范围/(°)	纬度范围/(°)	任务收益范围	卫星数量	卫星转换时间/s	观测时长范围/s	存储占用率/%	地面站数量	下传时长范围/s
G1~G6	[-50, 130]	[-20, 60]	[1, 100]	4	15	[15, 25]	1	4	[100, 130]

实例	CPLEX		SM_G3			GA			SA算法			TS算法			PN
实例	平均任务收益	平均计算时间/s	平均任务收益	GAP/%	平均计算时间/s	平均任务收益	GAP/%	平均计算时间/s	平均任务收益	GAP/%	平均计算时间/s	平均任务收益	GAP/%	平均计算时间/s	平均任务收益	GAP/%	平均计算时间/s
G1	5 602	81	5 506	1.72	4	5 493	1.95	42	5 492	1.97	45	5 488	2.04	39	5 497	1.87	4
G2	6 408	128	6 282	1.96	6	6 216	2.99	61	6 201	3.22	56	6 166	3.77	46	6 244	2.56	5
G3	7 054	207	6 896	2.24	6	6 804	3.55	78	6 801	3.58	79	6 757	4.21	54	6 826	3.23	5
G4	-	-	7 409	-	8	7 237	-	89	7 259	-	84	7 191	-	71	7 259	-	7
G5	-	-	7 901	-	10	7 661	-	108	7 696	-	97	7 613	-	83	7 636	-	7
G6	-	-	8 151	-	11	7 905	-	137	7 920	-	112	7 851	-	104	7 897	-	8

实例	SM_G1	SM_G2	SM_G3	SM_G4	SM_G5	SM_G6	平均任务收益差/%
G1	5 502	5 497	5 503	5 500	5 502	5 500	0.10
G2	6 292	6 300	6 303	6 293	6 296	6 295	0.17
G3	6 904	6 916	6 910	6 906	6 904	6 907	0.17
G4	7 388	7 389	7 398	7 384	7 380	7 390	0.24
G5	7 882	7 891	7 893	7 886	7 889	7 891	0.14
G6	8 124	8 139	8 147	8 136	8 135	8 144	0.28

[1]	孟麟芝, 孙小涓, 胡玉新, 高斌, 孙国庆, 牟文浩. 面向卫星在轨处理的强化学习任务调度算法[J]. 系统工程与电子技术, 2025, 47(6): 1917-1929.
[2]	郑康洁, 张新宇, 王伟菘, 刘震生. DQN与规则结合的智能船舶动态自主避障决策[J]. 系统工程与电子技术, 2025, 47(6): 1994-2001.
[3]	刘书含, 李彤, 李富强, 杨春刚. 意图态势双驱动的数据链抗干扰通信机制[J]. 系统工程与电子技术, 2025, 47(6): 2055-2064.
[4]	王雯, 赵凯南, 杨林, 杨雄军. 面向复杂场景的层次式任务规划方法[J]. 系统工程与电子技术, 2025, 47(4): 1255-1264.
[5]	熊威, 张栋, 任智, 杨书恒. 面向有人/无人机协同打击的智能决策方法研究[J]. 系统工程与电子技术, 2025, 47(4): 1285-1299.
[6]	马鹏, 蒋睿, 王斌, 徐盟飞, 侯长波. 基于隐式对手建模的策略重构抗智能干扰方法[J]. 系统工程与电子技术, 2025, 47(4): 1355-1363.
[7]	唐开强, 傅汇乔, 刘佳生, 邓归洲, 陈春林. 基于深度强化学习的带约束车辆路径分层优化研究[J]. 系统工程与电子技术, 2025, 47(3): 827-841.
[8]	陈夏瑢, 李际超, 陈刚, 刘鹏, 姜江. 基于异质网络的装备体系组合发展规划问题[J]. 系统工程与电子技术, 2025, 47(3): 855-861.
[9]	张庭瑜, 曾颖, 李楠, 黄洪钟. 基于深度强化学习的航天器功率-信号复合网络优化算法[J]. 系统工程与电子技术, 2024, 46(9): 3060-3069.
[10]	夏雨奇, 黄炎焱, 陈恰. 基于深度Q网络的无人车侦察路径规划[J]. 系统工程与电子技术, 2024, 46(9): 3070-3081.
[11]	杨志鹏, 陈子浩, 曾长, 林松, 毛金娣, 张凯. 复杂环境下的飞行器在线航路规划决策方法[J]. 系统工程与电子技术, 2024, 46(9): 3166-3175.
[12]	郭宏达, 娄静涛, 徐友春, 叶鹏, 李永乐, 陈晋生. 基于MADDPG的多无人车协同事件触发通信[J]. 系统工程与电子技术, 2024, 46(7): 2525-2533.
[13]	李阳阳, 罗俊仁, 张万鹏, 项凤涛. 多星协同观测遗传-演进双层任务规划算法[J]. 系统工程与电子技术, 2024, 46(6): 2044-2053.
[14]	张梦钰, 豆亚杰, 陈子夷, 姜江, 杨克巍, 葛冰峰. 深度强化学习及其在军事领域中的应用综述[J]. 系统工程与电子技术, 2024, 46(4): 1297-1308.
[15]	尹帅, 余建慧, 宋斌, 郭延宁, 李传江, 吕跃勇. 基于多种群混沌遗传算法的GEO目标服务任务规划[J]. 系统工程与电子技术, 2024, 46(3): 914-921.