基于路网感知的时空轨迹聚类

摘要

By researching the traditional trajectory clustering algorithm ,aiming at the disadvantages of the current algorithm ,an algorithm for road‐network aware spatial‐temporal trajectory clustering was proposed .The adjacent spatial‐temporal trajectories units were merged selectively by dividing the original trajectories based on real road‐network to ensure high continuity and high flow of the mobile objects .The motion paths of the mobile objects in certain time slice were found ,which was helpful in traffic planning and location base services .Comprehensive empirical studies on large synthetic data demonstrate the correctness and effi‐ciency of the NEASTT algorithm .%通过研究传统的轨迹聚类算法，针对现有算法中存在的不足，提出一种基于路网感知的时空轨迹聚类算法（NEASTT）。依据真实路网信息划分轨迹，在保证时空强连续、高流量的情况下，有选择地合并相邻时空聚类单元，得到移动对象在某聚集时间片内的高流量、强连续运动路径，可为交通规划及基于位置的服务等应用提供更有利的帮助。实验结果表明，该聚类方法能有效挖掘时空轨迹聚类，揭示在一定时间片段内移动对象运动的代表路径。