基于时变网格的对地观测卫星搜索海上船舶方法

doi:10.13229/j.cnki.jdxbgxb20210194

摘要/Abstract

摘要：

为提高对地观测卫星搜索海上船舶效能，提出了一种根据卫星对地观测幅宽和船舶最大航速划分区域时空维度的时变网格模型，将船舶航迹预测问题转化为时变网格转移概率预测问题，有效降低了多步预测的复杂度。改进了序列到序列（Seq2Seq）模型，通过对搜索区域内大量历史航迹的学习，实现了较高精确度的多步时变网格转移预测。设计了面向时变网格的卫星观测任务规划算法，以实际AIS数据和卫星信息开展了仿真实验，实验结果表明：基于深度学习的Seq2Seq模型多步预测时变网格具有较高的精确度，有效提升了对地观测卫星搜索海上船舶的效能。

关键词: 计算机应用, 时变网格, 船舶搜索, 对地观测卫星, 深度学习, 序列到序列模型

Abstract:

In order to improve the effectiveness， a time-varying grid model is proposed to divide the spatial and temporal dimensions of the region according to the satellite earth observation width and the maximum vessel speed. The vessel track prediction problem is transformed into a time-varying grid transition probability prediction problem， which effectively reduces the complexity of multi-step prediction. The multi-step time-varying grid transfer prediction with high accuracy is realized by the improved sequence to sequence（Seq2Seq） model， which has learned a large number of historical tracks in the research area. A satellite observation task planning algorithm for time-varying grids is designed and simulation experiments are carried out based on actual AIS data and satellite information. The results of experiments show that the Seq2Seq model based on deep learning has high accuracy in multi-step prediction of time-varying grids， which effectively improves the effectiveness of vessel search by earth observation satellites.

Key words: compute application, time-varying grid, vessel search, earth observation satellite, deep learning, sequence to sequence model

中图分类号:

TP301.6

胡丹,孟新. 基于时变网格的对地观测卫星搜索海上船舶方法[J]. 吉林大学学报(工学版), 2022, 52(8): 1896-1903.

Dan HU,Xin MENG. Vessel search method by earth observation satellite based on time⁃varying grid[J]. Journal of Jilin University(Engineering and Technology Edition), 2022, 52(8): 1896-1903.

图/表 12

图1

图2

图3

图4

图5

表1

表2

表3

表4

图6

表5

表6

参考文献 21

1	Paul E B, Carmine P, David A B F. Optimal search,location and tracking of surface maritime targets by a constellation of surveillance satellites[R]. Edinburgh:DSTO Information Sciences Laboratory, 2002.
2	王慧林,邱涤珊,马满好,等.基于先验信息的海洋移动目标卫星成像侦测任务规划[J].火力与指挥控制,2011, 36(3): 105-110.
	Wang Hui-lin, Qiu Di-shan, Ma Man-hao, et al. Research on mission-planning of satellite imaging reconnaissance for ocean moving targets based on the prior information[J]. Fire Control & Command Control, 2011, 36(3): 105-110.
3	陈杰,邢利菊. 面向海洋移动目标成像侦察方法研究[J]. 计算机与数字工程, 2014, 42(3): 395-398.
	Cheng Jie, Xing Li-ju. Imaging reconnaissance method facing ocean motion target[J]. Computer & Digital Engineering, 2014, 42(3): 395-398.
4	梅关林,冉晓旻,范亮,等. 面向移动目标的卫星传感器调度技术研究[J]. 信息工程大学学报, 2016, 17(5): 513-517.
	Mei Guan-lin, Ran Xiao-min, Fan Liang, et al. Research on satellite sensor scheduling technology for moving target[J]. Journal of Information Engineering University, 2016, 17(5): 513-517.
5	Li J F, Geng-xi Y Z, Yao F, et al.Using multiple satellites to search for maritime moving targets based on reinforcement learning[J].Journal of Donghua University(English Edition), 2016, 33(5): 749-754.
6	张海龙,夏维,胡笑旋,等. 面向多障碍物海面卫星搜索动目标方法[J]. 系统工程与电子技术, 2018, 40(10): 2256-2262.
	Zhang Hai-long, Xia Wei, Hu Xiao-xuan, et al. Method for moving targets search by satellites on multi-obstacle sea[J]. Systems Engineering and Electronics, 2018, 40(10): 2256-2262.
7	夏忠,张海龙,靳鹏.基于信息融合的多星搜索动目标问题[J].火力与指挥控制, 2020, 45(7): 20-25, 30.
	Xia Zhong, Zhang Hai-long, Jin Peng. Research on moving target problem of multi-satellite search based on information fusion[J]. Fire Control & Command Control, 2020, 45(7): 20-25, 30.
8	Qiao S, Shen D, Wang X, et al. A self-adaptive parameter selection trajectory prediction approach via hidden markov models[J]. IEEE Transactions on Intelligent Transportation Systems, 2015, 16(1):284-296.
9	Rong H, Teixeira A P, Soares C G. Ship trajectory uncertainty prediction based on a Gaussian process model[J]. Ocean Engineering, 2019, 182(15): 499-511.
10	Liang Y, Zhang H. Ship track prediction based on ais data and pso optimized LSTM network[J]. International Core Journal of Engineering, 2020, 6(5): 23-33.
11	胡玉可,夏维,胡笑旋,等.基于循环神经网络的船舶航迹预测[J]. 系统工程与电子技术, 2020, 42(4): 871-877.
	Hu Yu-ke, Xia Wei, Hu Xiao-xuan, et al. Vessel trajectory prediction based on recurrent neural network[J]. Systems Engineering and Electronics, 2020,42(4):871-877.
12	Duc-Duy N, Chan L V, Muhammad I A. Vessel trajectory prediction using sequence-to-sequence models over spatial grid[C]∥The 12th ACM International Conference, New Zealand, 2018: 258-261.
13	游兰,韩雪薇,何正伟,等.基于改进Seq2Seq的短时AIS轨迹序列预测模型[J]. 计算机科学, 2020, 47(9): 169-174.
	You Lan, Han Xue-wei, He Zheng-wei, et al. Improved sequence-to-sequence model for short-term vessel trajectory prediction using AIS data streams[J]. Computer Science, 2020, 47(9): 169-174.
14	夏永泉,黄海鹏,王兵,等.一种基于改进的无监督深度学习自编码方法[J]. 科技通报, 2018, 34(7):183-187.
	Xia Yong-quan, Huang Hai-peng, Wang Bing, et al. Self-encoding method based on improved unsupervised deep learning[J]. Bulletin of Science and Technology, 2018, 34(7): 183-187.
15	Sutskever I, Vinyals O, Le Q V. Sequence to Sequence Learning with Neural Networks[C/OL]∥[2021-03-15].
16	徐谦,李颖,王刚.基于深度学习图像语义分割的机器人环境感知[J].吉林大学学报:工学版, 2019, 49(1): 248-260.
	Xu Qian, Li Ying, Wang Gang. Robotic environment sensing based on semantic segmentation by deep learning[J]. Journal of Jilin University(Engineering and Technology Edition), 2019, 49(1): 248-260.
17	Sepp Hochreiter, Technische Universität München, Fakultät für Informatik, et al. Long short-term memory[J]. Neural Computation, 1997, 9(8): 1735-1780.
18	刘畅. 船舶自动识别系统(AIS)关键技术研究[D]. 大连: 大连海事大学通信与信息系统, 2013.
	Liu Chang. Study of key technology of automatic identification system(AIS)[D].Dalian: Communication and Information Systems, Dalian Maritime University, 2013.
19	伊恩,约书亚,亚伦.深度学习[M].北京: 人民邮电出版社, 2017.
20	赵宏伟,刘晓涵,张媛,等.基于关键点注意力和通道注意力的服装分类算法[J]. 吉林大学学报:工学版, 2020, 50(5): 1765-1770.
	Zhao Hong-wei, Liu Xiao-han, Zhang Yuan, et al. Clothing classification algorithm based on landmark attention and channel attention[J].Journal of Jilin University(Engineering and Technology Edition), 2020, 50(5): 1765-1770.
21	冯荣强,赵磊,杨勇,等.计及电价和Attention机制的LSTM短期负荷预测模型[J].科技通报, 2020, 36(11): 57-62, 68.
	Feng Rong-qiang, Zhao Lei, Yang Yong, et al. LSTM-Short-term load forecasting model considering electricity price and attention mechanism[J]. Bulletin of Science and Technology, 2020, 36(11): 57-62, 68.

相关文章 15

[1]	周丰丰,朱海洋. 基于三段式特征选择策略的脑电情感识别算法SEE[J]. 吉林大学学报(工学版), 2022, 52(8): 1834-1841.
[2]	白天,徐明蔚,刘思铭,张佶安,王喆. 基于深度神经网络的诉辩文本争议焦点识别[J]. 吉林大学学报(工学版), 2022, 52(8): 1872-1880.
[3]	李晓英,杨名,全睿,谭保华. 基于深度学习的不均衡文本分类方法[J]. 吉林大学学报(工学版), 2022, 52(8): 1889-1895.
[4]	申铉京,张雪峰,王玉,金玉波. 像素级卷积神经网络多聚焦图像融合算法[J]. 吉林大学学报(工学版), 2022, 52(8): 1857-1864.
[5]	曲福恒,丁天雨,陆洋,杨勇,胡雅婷. 基于邻域相似性的图像码字快速搜索算法[J]. 吉林大学学报(工学版), 2022, 52(8): 1865-1871.
[6]	赵宏伟,张健荣,朱隽平,李海. 基于对比自监督学习的图像分类框架[J]. 吉林大学学报(工学版), 2022, 52(8): 1850-1856.
[7]	秦贵和,黄俊锋,孙铭会. 基于双手键盘的虚拟现实文本输入[J]. 吉林大学学报(工学版), 2022, 52(8): 1881-1888.
[8]	王军,徐彦惠,李莉. 低能耗支持完整性验证的数据融合隐私保护方法[J]. 吉林大学学报(工学版), 2022, 52(7): 1657-1665.
[9]	周丰丰,张亦弛. 基于稀疏自编码器的无监督特征工程算法BioSAE[J]. 吉林大学学报(工学版), 2022, 52(7): 1645-1656.
[10]	高明华,杨璨. 基于改进卷积神经网络的交通目标检测方法[J]. 吉林大学学报(工学版), 2022, 52(6): 1353-1361.
[11]	康耀龙,冯丽露,张景安,陈富. 基于谱聚类的高维类别属性数据流离群点挖掘算法[J]. 吉林大学学报(工学版), 2022, 52(6): 1422-1427.
[12]	王文军,余银峰. 考虑数据稀疏的知识图谱缺失连接自动补全算法[J]. 吉林大学学报(工学版), 2022, 52(6): 1428-1433.
[13]	陈雪云,贝学宇,姚渠,金鑫. 基于G⁃UNet的多场景行人精确分割与检测[J]. 吉林大学学报(工学版), 2022, 52(4): 925-933.
[14]	方世敏. 基于频繁模式树的多来源数据选择性集成算法[J]. 吉林大学学报(工学版), 2022, 52(4): 885-890.
[15]	李大湘,陈梦思,刘颖. 基于STA⁃LSTM的自发微表情识别算法[J]. 吉林大学学报(工学版), 2022, 52(4): 897-909.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

卫星名称	卫星代号	轨道高度/km	倾角/（°）	降交点地方时	幅宽/km	侧摆能力/（°）
资源一号04星	ZY1-04	778	98.5000	10：30 AM	60	±32
资源三号01星	ZY3-01	506	97.4210	10：30 AM	51	±32
资源三号02星	ZY3-02	505	97.4210	10：30 AM	51	±32
高分一号卫星	GF-01	645	98.0506	10：30 AM	60	±35
高分二号卫星	GF-02	631	97.0908	10：30 AM	45	±35
实践九号A星	SJ9-A	645	97.9820	10：30 AM	30	±35

卫星代号	两行根数
ZY1-04	1 40336U 14079A 21041.78300343 -.00000110 00000-0 -23164-4 0 9997 2 40336 98.5363 117.3056 0001856 97.6028 262.5367 14.35423066323875
ZY3-01	1 38046U 12001A 21041.77982109 .00000761 00000-0 37025-4 0 9991 2 38046 97.3946 118.1644 0004704 78.3672 32.1382 15.21336706504859
ZY3-02	1 41556U 16033A 21041.88121295 .00000634 00000-0 31299-4 0 9999 2 41556 97.3402 117.4403 0000436 284.5074 203.7350 15.21367274261165
GF-01	1 39150U 13018A 21041.79204228 .00000042 00000-0 13123-4 0 9993 2 39150 97.8455 118.2646 0020503 70.0849 290.2565 14.76565123420188
GF-02	1 40118U 14049A 21041.74098995 .00000065 00000-0 15082-4 0 9992 2 40118 97.8174 119.3216 0009141 110.3054 249.9140 14.80653293350338
SJ9-A	1 38860U 12056A 21041.73090786 .00000043 00000-0 12358-4 0 9997 2 38860 97.6461 84.4101 0029174 174.9656 185.1845 14.79610380449670

参数名称	参数值
编码器输入特征维度	8
编码器隐层神经元数量	128
编码器隐层层数	4
编码器输出特征维度	32
解码器输入特征维度	32
解码器隐层神经元数量	128
解码器隐层层数	4
解码器输出特征维度	18
训练批次大小	32
训练轮数	4000

序号	窗口开始时间	窗口结束时间	过区域次数
1	2019-01-02 22∶53∶49	2019-01-03 02∶13∶44	11
2	2019-01-03 23∶11∶17	2019-01-04 01∶00∶53	10
3	2019-01-05 23∶41∶39	2019-01-06 01∶49∶13	10
4	2019-01-11 22∶48∶54	2019-01-12 02∶01∶12	11
5	2019-01-14 23∶29∶34	2019-01-15 02∶07∶02	8
6	2019-01-23 23∶17∶27	2019-01-24 01∶12∶38	8

序号	基于高斯分布船舶运动预测的搜索方法		基于时变网格的船舶搜索方法
序号	最早搜索到船舶时间延时/min	搜索到船舶次数	最早搜索到船舶时间延时/min	搜索到船舶次数
1	117.50	3	117.37	4
2	/	0	169.97	2
3	/	0	178.75	2
4	110.48	2	110.47	2
5	/	0	82.25	2
6	/	0	87.85	1