Origin-destination trips by purpose and time of day inferred from mobile phone data
Origin-destination trips by purpose and time of day inferred from mobile phone data
复制标题
DOI：
10.1016/j.trc.2015.02.018
复制
发表时间：
2015-09-01
影响因子：
8.3
通讯作者：
Gonzalez, Marta C.
中科院分区：
文献类型：
--
作者：
Alexander, Lauren;Jiang, Shan;Gonzalez, Marta C.
在这项工作中，我们提出了根据数百万匿名用户的三角移动电话记录来估计平均每日出发地-目的地行程的方法。这些记录首先被转换为用户在观察到的持续时间内从事活动的聚集位置。根据观察频率、一周中的某一天和一天中的时间，这些位置被推断为家庭、工作或其他位置，并且代表用户的出发地和目的地。由于这些位置的到达时间和持续时间反映的是观察到的（基于电话使用情况）而不是用户的真实到达时间和持续时间，因此我们使用美国主要城市旅行的调查数据来概率地推断出发时间。然后为每个用户在一天中的两次连续观察之间构建行程。这些出行乘以基于用户家庭人口普查区人口的扩展系数，再除以我们观察用户的天数，从而提取平均每日出行。按人口普查区对、一天中的时间和出行目的汇总个人的日常出行会产生出行矩阵，该矩阵构成了为交通规划和投资提供信息的大部分分析和建模的基础。所提出方法的适用性得到了对地方和国家调查中报告的旅行时间和空间分布的验证的支持。 (C) 2015 Elsevier Ltd. 保留所有权利。
In this work, we present methods to estimate average daily origin-destination trips from triangulated mobile phone records of millions of anonymized users. These records are first converted into clustered locations at which users engage in activities for an observed duration. These locations are inferred to be home, work, or other depending on observation frequency, day of week, and time of day, and represent a user's origins and destinations. Since the arrival time and duration at these locations reflect the observed (based on phone usage) rather than true arrival time and duration of a user, we probabilistically infer departure time using survey data on trips in major US cities. Trips are then constructed for each user between two consecutive observations in a day. These trips are multiplied by expansion factors based on the population of a user's home Census Tract and divided by the number of days on which we observed the user, distilling average daily trips. Aggregating individuals' daily trips by Census Tract pair, hour of the day, and trip purpose results in trip matrices that form the basis for much of the analysis and modeling that inform transportation planning and investments. The applicability of the proposed methodology is supported by validation against the temporal and spatial distributions of trips reported in local and national surveys. (C) 2015 Elsevier Ltd. All rights reserved.