基于AIS数据的海上电子围栏选址

Siting of maritime electronic fence based on AIS data

针对电子围栏在海洋交通管理应用中的选址粗放问题,利用船舶自动识别系统(AIS)数据聚类分析实现海上电子围栏的选址:通过数据清洗和特征抽取对AIS原始数据进行筛选,构建具有时序特征的属性数组;并采用改进的基于密度聚类(DBSCAN)算法对AIS数据进行聚类计算,挖掘分析出船舶的实际空间分布特征;然后给出海上电子围栏选址的具体方法和流程;最后以最小外接矩形方法获得研究区域电子围栏边界位置。实验结果表明,该方法获取的海上电子围栏在需求覆盖率、流速稳定性和靠近航道距离等方面的性能都有较大的提高,具有较为广阔的应用前景,可为港口的精细化、科学化管理提供技术支撑。

Aiming at the problem of rough site selection in the application of electronic fences in marine traffic management,the paper proposed a site selection method of marine electronic fences by using the clustering analysis of automatic ship identification system(AIS)data:the original AIS data were screened by data cleaning and feature extraction to construct an attribute array with time-series characteristics;and an improved density-based spatial clustering of applications with noise(DBSCAN)algorithm was used to cluster the AIS data to mine and analyze the actual spatial distribution characteristics of ships;then a specific method and its process for siting the maritime electronic fence were given;finally,the boundary location of the electronic fence in the study area was obtained by the minimum enclosing rectangle method.Experimental result showed that the performance of the maritime electronic fence obtained by the proposed method would be greatly improved in terms of demand coverage rate,flow speed stability and distance close to the channel,which could have a relatively broad application prospect and provide technical support for the refined and scientific management of the port.