文章编号： 1672-5328（2019）03-0011-08

吴子啸
(中国城市规划设计研究院，北京100037)

摘要： 利用手机数据推演居民出行特征是近年交通研究中的一个热点。尽管已有众多研究成果发 表，但现实应用中仍有诸多问题需要解决和改进。提出一种基于手机数据的出行链推演算法，通过 构造时空贪婪同化流程来处理手机数据的空间不确定性，并对传统聚类算法进行改进以提高活动地 点识别效率。通过个体实际数据验证了算法的有效性。结果显示，与其他方法相比，提出的算法能 够高效地锚固居住地、工作地等停留较长时间的出行端点，从而提高出行链推演的效率和准确性。 该算法适用于多天、手机基站定位和三角算法定位的混合位置数据，对现实数据有很好的适应性。

关键词： 大数据；手机数据；出行链；贪婪同化；聚类算法

中图分类号： U491.1+2

文献标识码：A

Travel Chain Estimation Based on Cell Phone Data

Wu Zixiao
(China Academy of Urban Planning & Design, Beijing 100037, China)

Abstract： Analyzing the characteristics of residents' travel demand using cell phone data is a popular topic among studies on transportation in recent years. Although there have been a great number of publications in this area, there are still numerous problems needing to be solved and improved in practical application. This paper proposes an algorithm for travel chain estimation based on cell phone data. This algorithm can effectively deal with the spatial uncertainty of cell phone data by developing the greedy assimilation process, and can increase the accuracy of activity location by improving traditional clustering algorithm. The efficiency of the algorithm is demonstrated through the application of actual data. The results show that compared with other methods, the proposed algorithm can efficiently identify travel origins/destinations with longer stay such as the residence and workplace. Accordingly, it improves the efficiency and accuracy of travel chain deduction. The algorithm is applicable to the multi-source location data from multiday positioning, positioning by cell phone base station, and triangle algorithm positioning, and is well adaptable to the actual data.

Keywords： big data; cell phone data; travel chain; greedy assimilation; clustering algorithms