为更好地应对船舶会遇局面混乱的问题，通过分析大量船舶AIS数据，按照让路船与直航船的标定，结合《规则》对会遇局面的3种划分，融合群体智慧，并利用支持向量机模型与决策树模型对全样本进行模拟，得到不同转向行为聚类特征，总结归纳出基于群体智慧的船舶避碰决策。最后运用综合风险评价模型对所得到的船舶避碰决策进行验证评估。结果表明，该船舶避碰决策研究在应对船舶会遇局面时能够克服《规则》对于传统会遇局面避碰行为指导中的风险衡量失真问题，研究成果可为船舶多船协同避碰提供一定参考。
In order to better cope with the confusion of the ship 's encounter situation, by analyzing a large number of ship AIS data, according to the calibration of the road ship and the straight ship, combined with the three divisions of the encounter situation by ' COLREGS ', integrated group wisdom, and the support vector machine model and the decision tree model are used to simulate the whole sample. The clustering characteristics of different steering behaviors are obtained, and the ship collision avoidance decision based on group wisdom is summarized. Finally, the comprehensive risk assessment model is used to verify and evaluate the obtained ship collision avoidance decision. The results show that the research on ship collision avoidance decision-making can overcome the problem of distortion of risk measurement in the guidance of ' COLREGS ' for traditional collision avoidance behavior in meeting situations. The research results can provide some reference for multi-ship cooperative collision avoidance.
2025,47(6): 41-48 收稿日期：2024-5-23
DOI：10.3404/j.issn.1672-7649.2025.06.007
分类号：U698.6
基金项目：广西科技计划项目（桂科AB23026132）
作者简介：王文灏（2000 – ），男，硕士研究生，研究方向为海上交通运输、交通安全与环境
参考文献：
[1] 吕红光, 裴天琪, 尹勇, 等. 智能船舶背景下《1972年国际海上避碰规则》的修正[J]. 上海海事大学学报, 2020, 41(4): 117-124.
[2] 张铎. 《1972年国际海上避碰规则公约》通过五十年回顾与展望[J]. 世界海运, 2022, 45(10): 1-5.
[3] CHAI T, WENG J, LI G. Estimation of vessel collision frequency in the Yangtze River estuary considering dynamic ship domains[J]. Journal of Marine Science and Technology, 2020, 25(3): 964-977.
[4] Maritime simulator based determination of minimum DCPA and TCPA in head-on ship-to-ship collision avoidance in confined waters[J] Transportmetrica A Transport Science, 2019, 15(2): 1124-1144.
[5] GAO M, SHI G Y, LIU J. Ship encounter azimuth map division based on automatic identification system data and support vector classification[J]. Ocean Engineering, 2020, 213: 107636.
[6] REN Y, MOU J, YAN Q, et al. Study on assessing dynamic risk of ship collision[C]//International Conference on Transportation Information and Safety,2011.
[7] YUAN X, ZHANG D, ZHANG J, et al. A novel real-time collision risk awareness method based on velocity obstacle considering uncertainties in ship dynamics[J]. Ocean Engineering, 2021, 220: 108436.
[8] ZHOU Y, DAAMEN W, VELLINGA T, et al. Ship behavior during encounters in ports and waterways based on AIS data: From theoretical definitions to empirical findings[J]. Ocean Engineering, 2023, 272: 113879.
[9] 羌鹏. AIS船舶时空大数据分析和可视化技术研究[D]. 大连: 大连海事大学, 2021.
[10] 黎法明, 胡甚平, 刘明明. 追越过程中的船舶避让行为反演分析[J]. 舰船科学技术, 2021, 43(21): 100-104.
LI F M, HU S P, LIU M M. Inversion analysis of ship avoidance behavior in the process of overtaking[J]. Ship Science and Technology, 2021, 43(21): 100-104.
[11] 胡佳颖, 刘克中, 吴晓烈, 等. 基于前景理论的船舶避碰决策优化方法[J]. 中国航海, 2020, 43(1): 18-23.
[12] KIM D, HIRAYAMA K, OKIMOTO T. Distributed stochastic search algorithm for multi-ship encounter situations[J]. Journal of Navigation, 2017, 70(4): 699-718.
[13] 徐武雄, 初秀民, 刘兴龙. 船舶行为可视分析关键技术研究[J]. 武汉理工大学学报(信息与管理工程版), 2015, 37(5): 451-454.
[14] RONG H, TEIXEIRA A P, GUEDES SOARES C. Ship collision avoidance behaviour recognition and analysis based on AIS data[J]. Ocean Engineering, 2022, 245: 110479.
[15] LI M, MOU J, CHEN L, et al. Comparison between the collision avoidance decision-making in theoretical research and navigation practices[J]. Ocean Engineering, 2021, 228: 108881.
[16] WANG T, YAN X, WANG Y, et al. A distributed model predictive control using virtual field force for multi-ship collision avoidance under COLREGs[C]//2017 4th International Conference on Transportation Information and Safety (ICTIS). Banff, AB, Canada: IEEE, 2017: 296-305[2024-05-08].
[17] 杜晓雨, 胡甚平, 李圣媛, 等. 船舶会遇过程碰撞风险演化随机Petri网模型[J]. 上海海事大学学报, 2023, 44(4): 25-32, 98.
[18] 张金奋, 游斌, 蔡明佑, 等. AIS数据驱动下的渡船横越场景识别与碰撞风险建模[J]. 中国航海, 2023, 46(2): 46-53, 59.
[19] 王耀伟, 刘原麟, 潘若禹, 等. 基于支持向量机的物流传送带场景下RFID识读性能研究[J]. 现代电子技术, 2024, 47(9): 150–156.
[20] 肖亮, 曾云. 基于决策树特征提取与RSM_LightGBM的涡扇发动机RUL预测[J]. 农业装备与车辆工程, 2024, 62(4): 132-136.
[21] 陈蜀喆, 康杰, 龚彪, 等. 基于黎曼球面投影变换的船舶会遇风险感知建模[J]. 武汉理工大学学报(交通科学与工程版), 2023: 1-9.
[22] 陈信强, 史飞翔, 王梓创, 等. 基于模糊逻辑方法的多船会遇安全态势评估[J]. 广西大学学报(自然科学版), 2022, 47(5): 1327-1336.
