基于改进多输出支持向量的船舶航迹预测

doi:10.12305/j.issn.1001-506X.2024.01.20

摘要/Abstract

摘要：

为保障智能船舶快速、安全、可靠地进行避碰工作, 提出了一种基于改进樽海鞘群算法(salp swarm algorithm, SSA)的多输出支持向量的船舶航迹预测模型。本文采用的多输出支持向量模型可以对船舶进行整体建模, 所构建的模型可以对船舶航迹状态进行多输出预测, 对于模型中存在的超参数采用改进的SSA进行寻优, 算法加入了自适应权重与离群象算法, 避免了算法早熟与高维易陷入局部最优的问题。最后，实验选取了实测数据对所提方法进行验证, 并与其他常见模型进行对比实验, 结果表明了所提方法的可行性与有效性。

关键词: 多输出支持向量回归, 樽海鞘群算法, 船舶航迹预测, 数据驱动

Abstract:

In order to ensure the rapid, safe and reliable collision avoidance of intelligent vessel, this paper proposes a vessel track prediction model based on the improved salp swarm algorithm (SSA) multi output support vector is proposed. The multi output support vector model used in this paper can model the vessel as a whole, and the vessel model built can predict the changes of vessel track status at the same time. For the super parameters in the model, the improved SSA is used for optimization. The algorithm adds the characteristics of adaptive weight and outlier algorithm, avoiding the problem of premature algorithm and local optimization that is easy to get stuck in high-dimensional. Finally, the proposed method is validated by measured data and compared with other common models. The results show that the proposed method is feasible and effective.

Key words: multi-output support vector regression (SVR), salp swarm algorithm (SSA), vessel trajectory prediction, data driven

中图分类号:

TP181

杨振亚, 张智, 尚晓兵, 曹择骏, 孙喆轩. 基于改进多输出支持向量的船舶航迹预测[J]. 系统工程与电子技术, 2023, 46(1): 173-181.

Zhenya YANG, Zhi ZHANG, Xiaobing SHANG, Zejun CAO, Zhexuan SUN. Vessel trajectory prediction based on improved multi-output support vector[J]. Systems Engineering and Electronics, 2023, 46(1): 173-181.

图/表 10

图1

图2

图3

表1

图4

图5

图6

图7

图8

表2

参考文献 30

1	WANG Y G , ZENG Y L . Research on maritime navigation perception requirements of intelligent ships[J]. Journal of Physics: Conference Series, 2022, 2356 (1): 96- 98.
2	ZHAI P Y , ZHANG Y J , WANG S B . Intelligent ship collision avoidance algorithm based on DDQN with prioritized experience replay under COLREGs[J]. Journal of Marine Science and Engineering, 2022, 10 (5): 45- 46.
3	谢朔. 基于天牛须优化的船舶运动建模与避碰方法研究[D]. 武汉: 武汉理工大学, 2020.
	XIE S. Research on ship motion modeling and collision avoidance method based on longicorn whisker optimization[D]. Wuhan: Wuhan University of Technology, 2020.
4	崔瑾娟. 基于遗传算法规划路径的船舶避碰系统[J]. 舰船科学技术, 2019, (12): 43- 45.
	CUI J J . Ship collision avoidance system based on genetic algorithms for path planning[J]. Ship Science and Technology, 2019, (12): 43- 45.
5	MURRAY B, PERERA L P. A data-driven approach to vessel trajectory predictionfor safeautonomous vessel operations[C]//Proc. of the International Conference on DigitalInformation Ma-nagement, 2018: 240-247.
6	PERERA L , OLIVEIRA P , SOARES C . 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
7	徐铁, 蔡奉君, 胡勤友, 等. 基于卡尔曼滤波算法船舶AIS轨迹估计研究[J]. 现代电子技术, 2014, 37 (5): 97- 100.
	XU T , CAI F J , HU Q Y , et al. Research on ship AIS trajectory estimation based on Kalman filter algorithm[J]. Modern Electronic Technology, 2014, 37 (5): 97- 100.
8	MAZZARELLA F, ARGUEDAS V F, VESPE M. Knowledge-based vessel position prediction using historical AIS data[C]//Proc. of the Workshop on Sensor Data Fusion: Trends, Solutions, Applications, 2015.
9	LIU J , GUO Y S . Vessel trajectory prediction model based on AIS sensor data and adaptive chaos differential evolution support vector regression (ACDE-SVR)[J]. Applied Sciences, 2019, 9 (15): 2983. doi: 10.3390/app9152983
10	MIZUNO N , KURODA M , OKAZAKI T , et al. Minimum time ship maneuvering method using neural network and nonlinear model predictive compensator[J]. Control Engineering Practice, 2007, 15 (6): 757- 765. doi: 10.1016/j.conengprac.2007.01.002
11	LIN R Q , HUGHES M , SMITH T . Prediction of ship steering capabilities with a fully nonlinear ship motion model. Part 1: maneuvering in calm water[J]. Journal of Marine Science and Technology, 2010, 15 (2): 131- 142. doi: 10.1007/s00773-010-0084-z
12	DASH A K , NAGARAJAN V , SHA O P . Bifurcation analysis of a high-speed twin-propeller twin-rudder ship maneuvering model in roll-coupling motion[J]. Nonlinear Dynamics, 2016, 83 (4): 2035- 2053. doi: 10.1007/s11071-015-2463-9
13	LE Q I , ZHENG Z . Trajectory prediction of vessels based on data mining and machine learning[J]. Journal of Digital Information Management, 2016, 14 (1): 192- 196.
14	许玲, 邱晗. 船舶静水航行水动力模型的简化及其应用[J]. 安徽工业大学学报: 自然科学版, 2014, 31 (2): 162- 166.
	XU L , QIU H . Simplification and application of hydrodynamic model for ships sailing in still water[J]. Journal of Anhui University of Technology: Natural Science Edition, 2014, 31 (2): 162- 166.
15	ABRAHA M S , RAISEEM G , BANIAS L G , et al. A robust and efficient stepwise regression method for building sparse polynomial chaos expansions[J]. Journal of Computational Physics, 2016, 332 (1): 461- 474.
16	ZHANG Y Y , WANG Z H , ZOU Z J . Black-box modeling of ship maneuvering motion based on multi-output nu-support vector regression with random excitation signal[J]. Ocean Engineering, 2022, 257, 111279. doi: 10.1016/j.oceaneng.2022.111279
17	ZHUANG C X , CHEN C . Research on autonomous route generation method based on AIS ship trajectory big data and improved LSTM algorithm[J]. Frontiers in Neurorobotics, 2022, 16, 1049343. doi: 10.3389/fnbot.2022.1049343
18	叶文哲. 基于机器学习和AIS数据的船舶轨迹预测模型的设计与实现[D]. 成都: 电子科技大学, 2022.
	YE W Z. Design and implementation of ship trajectory prediction model based on machine learning and AIS data[D] Chengdu: University of Electronic Science and Technology of China, 2022.
19	HOU X R , ZOU Z J , LIU C . Nonparametric identification of nonlinear ship roll motion by using the motion response in irregular waves[J]. Applied Ocean Research, 2018, 73, 88- 99. doi: 10.1016/j.apor.2018.02.004
20	KAWAN B, HAO W, LI G, et al. Data-driven modeling of ship motion prediction based on support vector regression[C]//Proc. of the 58th Conference on Simulation and Modelling, 2017.
21	BO Z, AIGUO S. Empirical mode decomposition based LSSVM for ship motion prediction[C]//Proc. of the International Symposium on Neural Networks, 2013.
22	RAN Y, SHI G Y, LI W F. Ship track prediction model based on automatic identification system data and bidirectional cyclic neural network[C]//Proc. of the 4th International Symposium on Traffic Transportation and Civil Architecture, 2021.
23	XU S , AN X , QIAO X D , et al. Multi-output least-squares support vector regression machines[J]. Pattern Recognition Letters, 2013, 34 (9): 1078- 1084. doi: 10.1016/j.patrec.2013.01.015
24	PARK J , JEONG J S , PARK Y . Ship trajectory prediction based on Bi-LSTM using spectral-clustered AIS data[J]. Journal of Marine Science and Engineering, 2021, 9 (9): 725- 730.
25	郑洪清, 谢聪, 周永权. 一种改进的樽海鞘群算法[J]. 广西科学, 2022, 29 (2): 287- 292.
	ZHENG H Q , XIE C , ZHOU Y Q . An improved coleoptile swarm algorithm[J]. Guangxi Science, 2022, 29 (2): 287- 292.
26	SANCHEZ-FERNANDEZ M , DE-PRADO-CUMPLIDO M , ARENAS-GARCIA J , et al. SVM multiregression for nonlinear channel estimation in multiple-input multiple-output systems[J]. IEEE Trans.on Signal Processing, 2004, 52 (8): 2298- 2307. doi: 10.1109/TSP.2004.831028
27	FERNANDO P C, GUSTAU C V, EMILIO S O. et al. Multi-dimensional function approximation and regression estimation[C]//Proc. of the International Conference on Artificial Neural Networks, 2002.
28	高天航, 徐力, 靳廉洁, 等. 考虑航艏向与数据变化差异的船舶轨迹预测[J]. 交通运输系统工程与信息, 2021, 21 (1): 90- 94.
	GAO T H, XU L, JIN L J, et al Ship trajectory prediction considering the difference between heading and data change[J]. Transportation System Engineering and Information, 2021, 21(1): 90-94.
29	黄纯颖, 曾庆敏, 陈玲红, 等. 基于粒子群支持向量回归优化的循环流化床床温预测分析[J]. 能源工程, 2022, 42 (3): 11- 17.
	HUANG C Y , ZEND Q M , CHEN L H , et al. Prediction analysis of circulating fluidized bed temperature based on particle swarm support vector regression optimization[J]. Energy Engineering, 2022, 42 (3): 11- 17.
30	周欣然. 基于最小二乘支持向量机的在线建模与控制方法研究[D]. 长沙: 湖南大学, 2012.
	ZHOU X R. Research on online modeling and control method based on least squares support vector machine[D]. Changsha: Hunan University, 2012.

Lat	Lon	Sog	Sin	Cos	时间/s
0.666 87	0.238 80	0.007 732	0.131 951	0.161 562	0.103 5
0.666 89	0.238 80	0.005 155	0.203 993	0.097 036	0.001 7
0.666 90	0.238 80	0.007 732	0.108 696	0.188 743	0.001 6

船舶轨迹模型	e_jMS		r_j		COV
船舶轨迹模型	最坏	最好	最坏	最好	最坏	最好
SSALSSVM	0.009 39	3.28E-06	0.971 87	0.999 18	3.91E-06	1.64E-07
PSOSVR	0.008 96	9.99E-06	0.854 28	0.983 45	1.75E-06	6.90E-07
GRIMLSSVM	0.004 52	5.50E-05	0.905 41	0.998 73	5.15E-08	1.93E-08
SSAMLSSVR	0.002 68	5.59E-06	0.990 43	0.999 87	9.83E-08	2.68E-10

[1]	汤安迪, 韩统, 徐登武, 周欢, 谢磊. 使用高斯分布估计策略的改进樽海鞘群算法[J]. 系统工程与电子技术, 2022, 44(7): 2229-2240.
[2]	黄嘉, 常思江. 基于数据驱动的攻击时间和攻击角度控制导引律[J]. 系统工程与电子技术, 2022, 44(10): 3213-3220.
[3]	邢志伟, 刘洪恩, 李彪, 罗谦, 文涛, 陈肇欣. 基于时空关联网络的机场机位运行过程建模[J]. 系统工程与电子技术, 2021, 43(3): 722-730.
[4]	孟晨, 杨华晖, 王成, 马征. 数据驱动的武器系统电子元部件级故障诊断研究综述[J]. 系统工程与电子技术, 2021, 43(2): 574-583.
[5]	阳榴, 朱卫纲, 吕守业, 马爽. 面向非协作多功能雷达的波形单元提取方法[J]. 系统工程与电子技术, 2021, 43(10): 2843-2850.
[6]	王鹏, 杨妹, 祝建成, 鞠儒生, 李革. 面向数字孪生的动态数据驱动建模与仿真方法[J]. 系统工程与电子技术, 2020, 42(12): 2779-2786.
[7]	张翔, 李革, 王鹏. 基于动态数据驱动的反潜战仿真系统目标探测设计[J]. 系统工程与电子技术, 2018, 40(11): 2591-.
[8]	葛承垄, 朱元昌, 邸彦强, 胡志伟. 装备平行仿真技术的基础理论问题[J]. 系统工程与电子技术, 2017, 39(5): 1169-1177.
[9]	徐建国, 李孟军, 姜江, 游翰霖. 数据驱动的技术创新网络模体分析[J]. 系统工程与电子技术, 2017, 39(5): 1072-1077.