Abstract: Trajectories play an important role in analyzing the behavior of moving objects.Many researches have been conducted that retrieve similar trajectories of moving objects in Euclidean space rather than in road network space.However，in real applications，most moving objects are located in road network space.In this paper，the properties of similar trajectories are investigated in road network space and a spatio-temporal representation scheme is proposed for modeling the trajectories of moving objects.This spatio-temporal representation scheme effectively converts trajectory from the road network space to the Euclidean space.For measuring similarity between two trajectories，a new POI-distance algorithm is proposed which enhances the existing distance algorithm by reducing the insignificant nodes of a trajectory.Theory and experimental results show that this method provides not only a practical method for searching for similar trajectories but also a clustering method for trajectories.

Key words: spatio-temporal database, trajectories distance, Points Of Interesting（POI）, trajectories clustering

摘要： 在分析移动对象行为时，移动对象轨迹因包含大量的信息而具有重要的作用。在实际应用中移动对象常受限于空间网络而无法利用现有欧氏空间中轨迹及其距离处理技术。分析了道路网络空间轨迹相似性性质，提出一种移动对象轨迹建模的时空表示方法，能有效地将轨迹从道路网络空间转化到欧氏空间；同时提出了一种基于兴趣点POI（Points Of Interesting）距离的轨迹间相似性测量方法，有效地对轨迹进行化简并减少轨迹中节点的数目，从而降低算法时间复杂度。该方法不仅可以用于搜索相似轨迹，还可方便地应用到轨迹聚类的相关工作中。

关键词: 时空数据库, 轨迹距离, 兴趣点（POI）, 轨迹聚类