出行轨迹演绎性时序聚类分割算法

摘要

When studying transportation mode with large-scale travel trajectory data,need to not only extract the origin and destination model from the trip,but also determine the transportation mode. However,a journey often contains a variety of transportation modes. How to extract the travel patterns and enhance the effect of the final transportation planning is the key problem of transportation research. Based on mobile intelligent terminal to collect travel paths and the point of the travel path of different transportation mode,put forward the apri-ority sequential clustering segmentation algorithm for segmentation of the travel path. The method is verified by experiment. Results show that the algorithm for separating different transportation period is of very high accuracy.%对于大规模出行轨迹数据进行出行方式研究时，除了要对行程中OD(起点-终点)进行提取，还要对其间交通方式进行判别。然而，一段行程中，往往包含多种交通方式，如何更精细地从中提取多种交通方式，提升最终交通规划效果，是目前出行方式研究的关键问题。在使用移动智能终端采集出行轨迹的基础上，对出行轨迹进行不同交通方式的转换点提取，最终提出了演绎性时序聚类分割算法进行出行轨迹的分段，并对其进行实验验证。结果表明算法对于分割不同交通方式段，达到了很高的精度。