Complex Network Approach for the Complexity and Periodicity in Traffic Time Series

Abstract

In this paper, we firstly use the traffic flow data collected from loop detectors on freeway and measure the complexity of data by Lempel-Ziv algorithm at different temporal scales. Considering each day as a cycle and each cycle as a single node, we then construct complex networks by using the distribution of density and its derivative. In addition, the networks are analyzed in terms of some statistical properties, such as average path length, clustering coefficient, density, and average degree. Finally, we use the correlation coefficient matrix, adjacent matrix and closeness to exploit the periodicity in weekdays and weekends of traffic flow data.