哈尔滨工程大学学报2018，Vol.39Issue(3)：428-432,5.DOI:10.11990/jheu.201609033
船舶AIS轨迹快速自适应谱聚类算法
Fast self-tuning spectral clustering algorithm for AIS ship trajectory
摘要
Abstract
To conduct fast clustering of automatic identification system(AIS)ship trajectory data,in this paper,we propose a fast self-tuning spectral clustering(FSSC)algorithm based on the Hausdorff distance.The trajectory data are pre-processed by the Douglas-Peucker(DP)algorithm,which preserves the trajectory characteristics.Based on the Hausdorff distance,trajectory similarity measurement function and similarity matrix that can automatically choose the scaling parameters are proposed,and a spectral clustering algorithm is used to cluster the ship trajectory.To veri-fy the proposed method,we selected the estuary of the Yangtze River as a case study and the results indicate that the FSSC can obtain the main route in the marine navigation area.The consumption of computer resources is small,and the calculation speed is much faster than the usual clustering method.The proposed algorithm can provide a reference for the identification of main ship routes and improve the efficiency of maritime traffic management.关键词
船舶自动识别系统/船舶轨迹/Douglas-Peucker算法/数据压缩/Hausdorff距离/谱聚类Key words
automatic identification system(AIS)/trajectory/Douglas-Peucker algorithm/data compression/Haus-dorff distance/spectral clustering分类
交通工程引用本文复制引用
牟军敏,陈鹏飞,贺益雄,张行健,朱剑峰,荣昊..船舶AIS轨迹快速自适应谱聚类算法[J].哈尔滨工程大学学报,2018,39(3):428-432,5.基金项目
国家自然科学基金面上项目(51579201) （51579201）
武汉理工大学研究生自主创新基金项目(2015-zy-109,175212005). （2015-zy-109,175212005）
