论文部分内容阅读
基于ECDIS、AIS以及支持向量机,引入交通冲突研究方法,提出船舶定线制警戒区水上交通冲突数据自动采集解决方案。方案利用支持向量机学习训练冲突样本数据,得到判别冲突严重程度SVM分类模型,并将SVM软件包和SVM分类模型嵌入到ECDIS,结合实时AIS数据建立水上交通冲突数据自动采集系统,应用该系统对长江口船舶定线制A警戒区进行72 h水上交通冲突数据采集。实例结果表明:警戒区72 h船舶流量为681艘次,水上交通冲突为327个;严重、中等和一般冲突数量分别占总数的10.7%、21.4%和67.9%。系统运行稳定,可实现水上交通冲突数据自动采集与处理。
Based on ECDIS, AIS and support vector machines, this paper introduces a research method of traffic conflict and proposes a solution to automatically collect the data of maritime traffic conflict in the area of ship routing warning. The program uses SVM to train the conflict sample data to get the SVM classification model to determine the severity of conflict. The SVM software package and SVM classification model are embedded in ECDIS, and the automatic acquisition system of water traffic conflict data is established with real-time AIS data. A 72-hour water traffic conflict data acquisition was conducted at the A warning zone of the Yangtze River estuary. The results of the example show that the traffic volume in 72 h of the warning zone is 681 and the number of water traffic conflicts is 327. The number of serious, medium and general conflicts respectively accounts for 10.7%, 21.4% and 67.9% of the total. The system is stable and can automatically collect and process water traffic conflict data.