Exploring Human Spatio-Temporal Travel Behavior Based on Cellular Network Data: A Case Study of Hangzhou, China

Abstract

Rapid urbanization has burdened urban infrastructures and traffic. Understanding the travel behavior is a potential solution to traffic problems. The traditional travel surveys are trapped in the cost and update frequency. The advent of ubiquitous cellular network data offers opportunities to uncover the travel behavior in a novel way. This study presents the implement of cellular network data on the travel behavior study. Specifically, we proposed the method to detect individual stay points in different time and spatial scales. The inbound and outbound travel flows of every urban district were estimated to reveal its travel characteristic. The results are reached based on the real-world data of Hangzhou, China. The distribution of detected homes is compared with census data and show a consistency of 0.88. About 61.1% of Hangzhou people commute less than 5 km, and urban periphery areas travel longer than urban center areas. The inbound-dominated and outbound-dominated sub-districts is characterized based on the proportion of inbound and outbound trips. The results proved the potential of deriving knowledge from cellular network to serve urban infrastructure planning and management.