基于轨迹聚类的热点路径分析方法

Hot route analysis method based on trajectory clustering

随着智能终端、移动定位、无线通信等技术的快速发展,在交通、物流等应用领域,大量受路网约束的轨迹数据得以收集。利用轨迹数据分析热点路径,可以在时空和语义特征不变的前提下反映移动对象的运动和行为模式。在提取道路交叉点的基础上,引入轨迹的停留点语义,并将两者共同作为特征点进行轨迹划分,在轨迹聚类的基础上进行子轨迹权重分析,从而获得语义更为完整且用户关注度更高的热点路径。实验表明了轨迹划分和热点路径分析方法的有效性。

With the development of smart device, mobile positioning and wireless communication technologies, numerous trajectory data constrained by road network can be collected from application fields such as transportation and logistics. Hot route analysis using trajectory data is beneficial for reflecting the motion and aetion patterns of moving objects without chan- ging the spatio-temporal and semantic properties. In this paper, the crossing points of road network and semantic feature stops of trajectory are extracted as characteristic points for trajectory partition. The hot routes with higher semantic, integrity and user attention degree can be acquired by analyzing the weight of sub-trajectories based on the result of trajectory cluste- ring. Experiments show that the trajectory partition and hot route analysis methods are effective.