Recognition and classification method in multi-ship encounter scenarios based on topological graph
-
摘要: 针对复杂通航水域多船会遇场景表征模型缺乏、船间干扰关系难以辨析等问题，提出了一种基于拓扑图的多船会遇场景辨识和分类方法；考虑船舶随时空动态变化特性，对船舶自动识别系统数据进行时间切片划分，获得了可供研究的距离数据；基于船间哈文森距离引入“查找-验证-调整”聚类算法组成拓扑图时间序列，并自动生成多船会遇场景代表拓扑图；通过SimGNN模型计算了不同会遇场景代表拓扑图相似性，实现了多船会遇场景相似性度量，使用K近邻分类器完成多船会遇场景分类，分析了不同拓扑图数量和不同船舶类型的会遇过程；使用宁波—舟山水域某一天(24 h)真实数据进行试验分析。研究结果表明：通过提出的多船会遇场景辨识算法，精准识别出水域内794个有效多船会遇场景，其中2船会遇场景占比最高，3船会遇场景次之，4船及其以上会遇场景相对较少，该结果和船舶交通管理人员认知一致；多数场景持续时长维持在1 000 s内，生成拓扑图数量保持在100内，数据分布趋势较为近似；同一会遇场景内船舶数量波动较小，验证了所提出辨识算法的稳定性；使用了分类算法后，处于同分类的不同持续时长的场景间船舶类型和代表拓扑图具有明显相似性，不同分类间的场景在变化过程、持续时长、船舶类型和代表拓扑图上具有显著差别。Abstract: Aiming at the problems such as the lack of characterization models, and the challenge of distinguishing between ship interactions for multi-ship encounter scenarios in complex navigable waters, a recognition and classification method was proposed based on topological graph for multi-ship encounter scenarios. Considering the dynamic characteristics of ships in time and space, automatic identification system data was divided by time slices to obtain distance data suitable for research. Based on inter-ship Haversine distances, the find-verify-and-fix clustering algorithm was applied to construct time series of topological graphs, and the representative topological graphs were automatically generated in multi-ship encounter scenarios. The similarity between different representative topological graphs for different encounter scenarios was calculated by SimGNN model, and the similarity measurement of multi-ship encounter scenarios was realized. A K-nearest neighbors classifier was employed for multi-ship encounter scenario classification, and the encounter processes of different topological graph number and various ship types were analyzed. Experimental analysis was conducted using real data from a specific day (24 hours) in Ningbo-Zhoushan water area. Research results indicate that the proposed recognition algorithm for multi-ship encounter scenarios accurately identifies 794 valid multi-ship encounter scenarios in the water area, with two-ship encounter scenarios being the most common, followed by three-ship encounter scenarios, and relatively few occurrences of four-or-more-ship encounter scenarios. This result aligns with the perception of vessel traffic service personnel. Most scenarios have a duration of less than 1 000 s, with the numbers of generated topological graphs remaining below 100, indicating a relatively similar trend in data distribution. In the same encounter scenario, there is little fluctuation in ship number, demonstrating the stability of the proposed recognition algorithm. After using the classification algorithm, there is an obvious similarity in ship types and representative topological graphs between scenarios of different durations in the same category. Scenarios of various categories are significantly different in the evolution process, duration, ship type, and representative topological graph.
-
Key words:
- waterway transportation /
- multi-ship encounter scenario /
- FVF model /
- SimGNN model /
- AIS data /
- topological graph
-
图 9 3船会遇场景内GI和GU相似度分布
Figure 9. GI and GU similarity distributions in 3-ship encounter scenarios
图 10 7船会遇场景内GI和GU相似度分布
Figure 10. GI and GU similarity distributions in 7-ship encounter scenarios
图 16 最大、最小参与船舶数量差分布
Figure 16. Difference distribution between maximum and minimum numbers of participating ships
图 18 3船同分类会遇场景代表拓扑图实例
Figure 18. 3-ship encounter scenarios representative topological graphs example in same classification
图 19 5船不同分类会遇场景代表拓扑图实例
Figure 19. 5-ship encounter scenarios representative topological graphs example in different classifications
表 1 SimGNN模型参数
Table 1. Parameters of SimGNN model
参数名称 参数值 训练集 AIDS 训练样本数 700 瓶颈层神经元数 16 卷积层1神经元数 128 卷积层2神经元数 64 卷积层3神经元数 32 张量层神经元数 16 直方图箱数 16 迭代次数 50 学习率 0.001 丢弃率 0.5 权重衰减 0.001 
下载: 导出CSV
-
[1] CAI Ming-you, ZHANG Jin-fen, ZHANG Di, et al. Collision risk analysis on ferry ships in Jiangsu section of the Yangtze River based on AIS data[J]. Reliability Engineering and System Safety, 2021, 215: 107901. doi: 10.1016/j.ress.2021.107901 [2] 文元桥, 杜磊, 黄亚敏, 等. 水上交通流宏观复杂度建模与仿真[J]. 系统仿真学报, 2017, 29(4): 826-831.WEN Yuan-qiao, DU Lei, HUANG Ya-min, et al. Modeling and simulation of marine traffic macroscopic flow complexity[J]. Journal of System Simulation, 2017, 29(4): 826-831. (in Chinese) [3] SZLAPCZYNSKI R, SZLAPCZYNSKA J. An analysis of domain-based ship collision risk parameters[J]. Ocean Engineering, 2016, 126: 47-56. doi: 10.1016/j.oceaneng.2016.08.030 [4] 冮龙晖, 郑中义, 齐乐. AIS数据中船舶会遇信息的提取方法[J]. 中国科技论文, 2017, 12(7): 802-805. doi: 10.3969/j.issn.2095-2783.2017.07.016GANG Long-hui, ZHENG Zhong-yi, QI Le. Extraction of ship-encounter information from AIS data[J]. China Sciencepaper, 2017, 12(7): 802-805. (in Chinese) doi: 10.3969/j.issn.2095-2783.2017.07.016 [5] 马杰, 刘琪, 张春玮, 等. 基于AIS的数据时空分析及船舶会遇态势提取方法[J]. 中国安全科学学报, 2019, 29(5): 111-116.MA Jie, LIU Qi, ZHANG Chun-wei, et al. A method for extracting ship encounter situation based on spatio-temporal analysis of AIS data[J]. China Safety Science Journal, 2019, 29(5): 111-116. (in Chinese) [6] 马杰, 李文楷, 张春玮, 等. 基于AIS数据的交汇水域船舶会遇态势辨识[J]. 中国航海, 2021, 44(1): 68-74.MA Jie, LI Wen-kai, ZHANG Chun-wei, et al. Ship encounter situation recognition by processing AIS data from traffic intersection waters[J]. Navigation of China, 2021, 44(1): 68-74. (in Chinese) [7] WANG Lun-wei, CHEN Guo-cheng, LIU Jiong-jiong, et al. Multi-ship collision avoidance decision-making model considering operation conflict under COLREGs[C]//IEEE. 6th International Conference on Transportation Information and Safety: New Infrastructure Construction for Better Transportation. New York: IEEE, 2021: 429-435. [8] CHO Y, HAN J, KIM J. Efficient COLREG-compliant collision avoidance in multi-ship encounter situations[J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(3): 1899-1911. doi: 10.1109/TITS.2020.3029279 [9] WANG Shao-bo, ZHANG Ying-jun, LI Lian-bo. A collision avoidance decision-making system for autonomous ship based on modified velocity obstacle method[J]. Ocean Engineering, 2020, 215: 107910. doi: 10.1016/j.oceaneng.2020.107910 [10] CHEN Peng, SHI Guo-you, LIU Shuang, et al. Pattern knowledge discovery of ship collision avoidance based on AIS data analysis[J]. International Journal of Performability Engineering, 2018, 14(10): 2449-2457. [11] XIN Xu-ri, LIU Ke-zhong, LI Huan-huan, et al. Maritime traffic partitioning: an adaptive semi-supervised spectral regularization approach for leveraging multi-graph evolutionary traffic interactions[J]. Transportation Research Part C: Emerging Technologies, 2024, 164: 104670. doi: 10.1016/j.trc.2024.104670 [12] GAO Da-wei, ZHU Yong-sheng, YAN Ke, et al. Deep learning-based framework for regional risk assessment in a multi-ship encounter situation based on the transformer network[J]. Reliability Engineering and System Safety, 2024, 241: 109636. doi: 10.1016/j.ress.2023.109636 [13] LI Y P, LIU Z J, KAI J S. Study on complexity model and clustering method of ship to ship encountering risk[J]. Journal of Marine Science and Technology, 2019, 27(2): 8. [14] 胡文博, 黄蔚, 胡国超. 基于OPTICS聚类和关联分析的轨迹伴随模式分析[J]. 计算机与现代化, 2017(12): 82-87.HU Wen-bo, HUANG Wei, HU Guo-chao. Trajectory adjoint pattern analysis based on OPTICS clustering and association analysis[J]. Computer and Modernization, 2017(12): 82-87. (in Chinese) [15] BRCKO T, LUIN B. A decision support system using fuzzy logic for collision avoidance in multi-vessel situations at sea[J]. Journal of Marine Science and Engineering, 2023, 11(9): 1819. doi: 10.3390/jmse11091819 [16] 甄荣, RIVEIRO M, 金永兴. 一种基于DBSCAN的船舶会遇实时识别方法[J]. 上海海事大学学报, 2018, 39(1): 1-5.ZHEN Rong, RIVEIRO M, JIN Yong-xing. A real-time identification method to ship encounter based on DBSCAN[J]. Journal of Shanghai Maritime University, 2018, 39(1): 1-5. (in Chinese) [17] ZHU Ji-xiang, GAO Miao, ZHANG An-min, et al. Multi-ship encounter situation identification and analysis based on AIS data and graph complex network theory[J]. Journal of Marine Science and Engineering, 2022, 10: 1536. doi: 10.3390/jmse10101536 [18] 东昉, 徐建红, 苏开文. 船舶会遇态势的判断[J]. 航海技术, 2007(1): 2-5.DONG Fang, XU Jian-hong, SU Kai-wen. Judgment of ship encounter situation[J]. Marine Technology, 2007(1): 2-5. (in Chinese) [19] 郑中义, 吴兆麟. 多船会遇避碰决策研究——Ⅰ多船会遇局面定义、划分及属于各类判断[J]. 航海技术, 2000(4): 9-12.ZHENG Zhong-yi, WU Zhao-lin. Study on decision-making of collision avoidance in multi-ship encounter—Ⅰ definition, division and judgment of multi-ship encounter situation[J]. The Technology of Navigation, 2000(4): 9-12. (in Chinese) [20] 吴春杰, 李加庆. 浅谈多船会遇局面的界定与划分[J]. 武汉船舶职业技术学院学报, 2009, 8(4): 23-26.WU Chun-jie, LI Jia-qing. Analysis of multiple encounter situations' definition and classification[J]. Journal of Wuhan Vocational and Technical Shipping College, 2009, 8(4): 23-26. (in Chinese) [21] LI Tian-yu, XU Hui-qi, ZENG Wei-gui. Ship classification method for massive AIS trajectories based on GNN[J]. Journal of Physics: Conference Series, 2021, 2025(1): 012024. [22] ZAKI A, ATTIA M, HEGAZY D, et al. Comprehensive survey on dynamic graph models[J]. International Journal of Advanced Computer Science and Applications, 2016, 7(2): 573-582. [23] REN Cheng-hui, LO E, KAO B, et al. On querying historical evolving graph sequences[J]. Proceedings of the VLDB Endowment, 2011, 4(11): 726-737. [24] GILLIOZ A, RIESEN K. Graph-based pattern recognition on spectral reduced graphs[J]. Pattern Recognition, 2023, 144: 109859. [25] BUNKE H, SHEARER K. A graph distance metric based on the maximal common subgraph[J]. Pattern Recognition Letters, 1998, 19(3/4): 255-259. [26] BAI Yun-sheng, DING Hao, BIAN Song, et al. SimGNN: a neural network approach to fast graph similarity computation[C]//ACM. 12th ACM International Conference on Web Search and Data Mining. New York: ACM, 2019: 384-392. [27] REN Wei-kai, JIN Ning-de, OUYANG Lei. Phase space graph convolutional network for chaotic time series learning[J]. IEEE Transactions on Industrial Informatics, 2024, 20(5): 7576-7584. [28] RICHARD S, CHEN Dan-qi, CHRISTOPHER D, et al. Reasoning with neural tensor networks for knowledge base completion[C]//NeurIPS. 27th Annual Conference on Neural Information Processing Systems. San Diego: NeurIPS, 2013: 926-935. [29] XIN Xu-ri, LIU Ke-zhong, LOUGHNEY S, et al. Graph-based ship traffic partitioning for intelligent maritime surveillance in complex port waters[J]. Expert Systems with Applications, 2023, 231(11): 120825. [30] HU Liang, NAEEM W, RAJABALLY E, et al. A multiobjective optimization approach for COLREGs-compliant path planning of autonomous surface vehicles verified on networked bridge simulators[J]. IEEE Transactions on Intelligent Transportation Systems, 2020, 21(3): 1167-1179. -
点击查看大图
图(19) / 表(1)
计量
- 文章访问数: 3
- HTML全文浏览量: 1
- PDF下载量: 2
- 被引次数: 0
