基于循环神经网络的船舶航迹预测

doi:10.3969/j.issn.1001-506X.2020.04.18

系统工程与电子技术 ›› 2020, Vol. 42 ›› Issue (4): 871-877.doi: 10.3969/j.issn.1001-506X.2020.04.18

基于循环神经网络的船舶航迹预测

胡玉可^1,²(), 夏维^1,²(), 胡笑旋^1,²(), 孙海权^1,²(), 王云辉^1,²()

1. 合肥工业大学管理学院, 安徽合肥 230009
2. 过程优化与智能决策教育部重点实验室, 安徽合肥 230009

收稿日期:2019-05-05 出版日期:2020-03-28 发布日期:2020-03-28
作者简介:胡玉可(1995-),女,硕士研究生,主要研究方向为决策理论与方法。E-mail:1169042802@qq.com|夏维(1983-),男,讲师,博士,主要研究方向为智能决策。E-mail:xiawei@hfut.edu.cn|胡笑旋(1978-),男,教授,博士,主要研究方向为空间信息网络任务规划与资源调度。E-mail:xiaoxuanhu@hfut.edu.cn|孙海权(1993-),男,博士研究生,主要研究方向为卫星智能规划。E-mail:sunhaiquan2015@163.com|王云辉(1986-),男,博士研究生,主要研究方向为卫星任务规划。E-mail:97069570@qq.com
基金资助:
国家自然科学基金(71671059);国家自然科学基金(71521001);国家自然科学基金(71871079)

Vessel trajectory prediction based on recurrent neural network

Yuke HU^1,²(), Wei XIA^1,²(), Xiaoxuan HU^1,²(), Haiquan SUN^1,²(), Yunhui WANG^1,²()

1. School of Management, Hefei University of Technology, Hefei 230009, China
2. Key Laboratory of Process Optimization and Intelligent Decision-Making, Ministry of Education, Hefei 230009, China

Received:2019-05-05 Online:2020-03-28 Published:2020-03-28
Supported by:
国家自然科学基金(71671059);国家自然科学基金(71521001);国家自然科学基金(71871079)

摘要/Abstract

摘要：

在海事搜救、海关缉私等应用中,对目标船舶进行航迹预测是一个关键问题。为提高预测的精度和效率,提出了一种基于循环神经网络的船舶航迹预测方法,该方法包含数据预处理和神经网络预测两个部分。在数据预处理中,设计了一种基于对称分段路径距离的数据预处理方法,消除了大量冗余数据及噪声的影响;在神经网络预测中,构建了基于门控循环单元的循环神经网络模型,实现船舶位置信息精准且高效的预测。通过大量船舶自动识别系统数据进行了对比实验,实验结果证明了方法的实用性和有效性。

关键词: 航迹预测, 船舶自动识别系统, 对称分段路径距离, 门控循环单元

Abstract:

In maritime search and rescue, customs anti-smuggling and other scenarios, it is often necessary to forecast vessels' trajectory. In order to improve the accuracy and efficiency of the prediction, a method for vessel trajectory prediction based on recurrent neural network is proposed. The method includes data preprocessing and neural network prediction. In data preprocessing, a data preprocessing method based on symmetric segmented-path distance is designed to eliminate the influence of a large number of redundant data and noise. In the prediction of neural network, the model of recurrent neural network with gated recurrent unit as the core is constructed to realize the accurate and efficient prediction of vessels'position information. Comparative experiment is made through a large number of data from the automatic identification system, and experiment results prove that the proposed method is practical and effective.

Key words: trajectory prediction, automatic identification system, symmetrized segment-path distance, gated recurrent unit

中图分类号:

TP183

胡玉可, 夏维, 胡笑旋, 孙海权, 王云辉. 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.

Yuke HU, Wei XIA, Xiaoxuan HU, Haiquan SUN, Yunhui WANG. Vessel trajectory prediction based on recurrent neural network[J]. Systems Engineering and Electronics, 2020, 42(4): 871-877.

图/表 13

图1

图2

图3

图4

表1

图5

表2

图6

图7

图8

表3

表4

表5

参考文献 28

1	GUO S, LIU C, GUO Z, et al. Trajectory prediction for ocean vessels base on K-order multivariate Markov chain[C]//Proc.of the International Conference on Wireless Algorithms, Systems and Applications, 2018, 140-150.
2	乔少杰, 金琨, 韩楠, 等. 一种基于高斯混合模型的轨迹预测算法[J]. 软件学报, 2015, 26 (5): 1048- 1063.
	QIAO S J , JIN K , HAN N , et al. Trajectory prediction algorithm based on Gaussian mixture model[J]. Journal of Software, 2015, 26 (5): 1048- 1063.
3	HEXEBERG S, FLATEN A L, BREKKE E F. AIS-based vessel trajectory prediction[C]//Proc.of the 20th International Conference on Information Fusion, 2017: 1-8.
4	PERERA L P , OLIVEIRA P , SOARES C G . Maritime traffic monitoring based on vessel detection, tracking, state estimation, and trajectory prediction[J]. IEEE Trans.on Intelligent Transportation Systems, 2012, 13 (3): 1188- 1200. doi: 10.1109/TITS.2012.2187282
5	PERERA L P, SOARES C G. Ocean vessel trajectory estimation and prediction based on extended kalman filter[C]//Proc.of the 2nd International Conference on Adaptive and Self-Adaptive Systems and Applications, 2010: 14-20.
6	彭秀艳, 门志国, 刘长德. 基于Kalman滤波算法的Volterra级数核估计及其应用[J]. 系统工程与电子技术, 2010, 32 (11): 2431- 2435.
	PENG X Y , MEN Z G , LIU C D . Volterra series kernel estimation based on Kalman filtering algorithm and its application[J]. Systems Engineering and Electronics Technology, 2010, 32 (11): 2431- 2435.
7	NGUYEN D D, LE V C, ALI M I. Vessel trajectory prediction using sequence-to-sequence models over spatial grid[C]//Proc.of the 12th ACM International Conference on Distributed and Event-based Systems, 2018: 258-261.
8	ZHU F, MARITIME D. Mining ship spatial trajectory patterns from AIS database for maritime surveillance[C]//Proc.of the 2nd International Conference on Emergency Management & Management Sciences, 2011: 772-775.
9	DOBRKOVIC A , IACOB M E , HILLEGERSBERG J V , et al. Towards an approach for long term AIS-based prediction of vessel arrival times[M]. Logistics and Supply Chain Innovation. Switzerland: Springer, 2016: 281- 294.
10	WIJAYA W M, NAKAMURA Y. Predicting ship behavior navigating through heavily trafficked fairways by analyzing AIS data on apache Hbase[C]//Proc.of the 1st International Symposium on Computing and Networking, 2014: 220-226.
11	任文娟, 周志鑫, 吕守业, 等. 基于AIS信息的舰船位置标校方法[J]. 系统工程与电子技术, 2016, 38 (10): 2381- 2388. doi: 10.3969/j.issn.1001-506X.2016.10.23
	REN W J , ZHOU Z X , LYU S Y , et al. Ship position calibration method based on AIS information[J]. Systems Engineering and electronics, 2016, 38 (10): 2381- 2388. doi: 10.3969/j.issn.1001-506X.2016.10.23
12	ZISSIS D , XIDIAS E K , LEKKAS D . Real-time vessel behavior prediction[J]. Evolving Systems, 2016, 7 (1): 29- 40. doi: 10.1007/s12530-015-9133-5
13	LEI P R , SU J , PENG W C , et al. A framework of moving behavior modeling in the maritime surveillance[J]. Journal of Chung Cheng Institute of Technology, 2011, 40 (2): 33- 42.
14	PALLOTTA G , VESPE M , BRYAN K . Vessel pattern knowledge discovery from AIS data: a framework for anomaly detection and route prediction[J]. Entropy, 2013, 15 (6): 2218- 2245.
15	VESPE M, VISENTINI I, BRYAN K, et al. Unsupervised learning of maritime traffic patterns for anomaly detection[C]//Proc.of the 9th Data Fusion & Target Tracking Conference, 2012: 110-114.
16	MAZZARELLA F , VESPE M , SANTAMARIA C . SAR ship detection and self-reporting data fusion based on traffic know-ledge[J]. IEEE Geoscience and Remote Sensing Letters, 2015, 12 (8): 1685- 1689. doi: 10.1109/LGRS.2015.2419371
17	MAZZARELLA F, ARGUEDAS V F, VESPE M. Knowledge-based vessel position prediction using historical AIS data[C]//Proc.of the Sensor Data Fusion: Trends, Solutions, Applications, 2015: 1-6.
18	权波, 杨博辰, 胡可奇, 等. 基于LSTM的船舶航迹预测模型[J]. 计算机科学, 2018, 45 (S2): 136- 141.
	QUAN B , YANG B C , HU K Q , et al. Ship track prediction model based on LSTM[J]. Computer Science, 2018, 45 (S2): 136- 141.
19	VALSAMIS A , TSERPES K , ZISSIS D , et al. Employing traditional machine learning algorithms for big data streams analysis: the case of object trajectory prediction[J]. Journal of Systems and Software, 2017, 127, 249- 257. doi: 10.1016/j.jss.2016.06.016
20	钟颖, 汪秉文. 基于遗传算法的BP神经网络时间序列预测模型[J]. 系统工程与电子技术, 2002, 24 (4): 9- 11. doi: 10.3321/j.issn:1001-506X.2002.04.003
	ZHONG Y , WANG B W . BP neural network time series prediction model based on genetic algorithm[J]. Systems Engineering and electronics, 2002, 24 (4): 9- 11. doi: 10.3321/j.issn:1001-506X.2002.04.003
21	BESSE P , GUILLOUET B , LOUBES J M , et al. Review and perspective for distance based trajectory clustering[J]. Computer Science, 2015, 47 (2): 169- 179.
22	CHEN J, WANG R, LIU L, et al. Clustering of trajectories based on Hausdorff distance[C]//Proc.of the International Conference on Electronics, Communications and Control, 2011: 1940-1944.
23	BENGIO Y , SIMARD P , FRASCONI P . Learning long-term dependencies with gradient descent is difficult[J]. IEEE Trans.on Neural Networks, 1994, 5 (2): 157- 166.
24	HINTON G E, SRIVASTAVA N, KRIZHEVSKY A, et al. Improving neural networks by preventing co-adaptation of feature detectors[J]. arXiv Preprint arXiv: 1207.0580, 2012.
25	KRIZHEVSKY A, SUTSKEVER I, HINTON G E. Imagenet classification with deep convolutional neural networks[C]//Proc.of the Advances in Neural Information Processing Systems, 2012: 1097-1105.
26	HEATON J . Programming neural networks with encog3 in C#[M]. 2nd ed Louis: Heaton Research, 2011.
27	KINGMA D P, BA J. Adam: a method for stochastic optimization[C]//Proc.of the 3rd International Conference for Learning Representations, arXiv: 1412.6980, 2014.
28	CHO K , VAN MERRIENBOER B , BAHDANAU D , et al. On the properties of neural machine translation: encoder-decoder approaches[J]. Computer Science, 2014.

处理器	Intel(R) Core(TM) i5-7500 3.40 GHz
内存	8.00 GB
系统类型	Windows 10操作系统(64位)
程序设计语言	Python 3.6.3
编程环境	Visual Studio Code
模型框架构建	Python的Tensorflow和Keras实现

模型	模型参数	数据组	耗时/s	预测MSE
SSPD 和 GRU	神经元数量=100 样本训练次数=2 批处理数量=32	数据1	102.41	0.001 76
		数据2	110.35	0.004 05
		数据3	118.64	0.006 13
		数据4	103.02	0.003 86
		数据5	121.63	0.006 12
		数据6	107.87	0.002 21
		数据7	104.67	0.001 35
		数据8	123.65	0.005 96
		数据9	110.83	0.002 61
		数据10	107.24	0.003 38

样本训练次数/次	计算耗时/s	MSE
1	53.069	0.004 321
2	103.808	0.003 585
3	154.889	0.002 472
4	204.947	0.002 069
5	259.472	0.001 662
6	314.956	0.001 566
7	366.846	0.001 364
8	420.121	0.001 399
9	478.946	0.001 249
10	524.858	0.001 555

模型	度量指标	批处理数量/个
模型	度量指标	16	24	32
GRU	误差(MSE)	0.002 768	0.002 728	0.003 585
GRU	计算耗时/s	142.932	128.941	103.808

模型	度量指标	神经元数量/个
模型	度量指标	40	60	80	100	120	140	160
GRU	MSE	0.004 489	0.004 239	0.003 885	0.003 585	0.003 871	0.003 494	0.004 841
GRU	计算耗时/s	79.350	94.883	113.567	147.835	169.937	208.444	272.471

基于循环神经网络的船舶航迹预测

Vessel trajectory prediction based on recurrent neural network

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 13

参考文献 28

相关文章 3

编辑推荐

Metrics

本文评价

[1]	马泽煊, 李进, 路艳丽, 陈晨. 融合WaveNet和BiGRU的网络入侵检测方法[J]. 系统工程与电子技术, 2022, 44(8): 2652-2660.
[2]	马钰棠, 孙鹏, 张杰勇, 王鹏, 闫云飞, 赵亮. 样本不平衡下的空中群组意图识别方法[J]. 系统工程与电子技术, 2022, 44(12): 3747-3755.
[3]	张宏鹏, 黄长强, 轩永波, 唐上钦. 用门控循环单元实时预测空战飞行轨迹[J]. 系统工程与电子技术, 2020, 42(11): 2546-2552.