Spread of Traffic Congestion Detected by Traffic Sensor Data Mining in Naha City

Abstract

Road traffic conditions in urban areas are complicated by the daily, weekly, seasonally, weather-induced traffic demand fluctuations, various origin–destination patterns of traffic demand, effects caused by the control of traffic signals installed in these areas, and so on. Therefore, it is not easy to quantitatively analyze typical traffic congestion patterns that are represented by the time and place of occurrence, the process of propagation and diminution, duration time, and many others.

This study analyzed and visualized the spread patterns of traffic congestion in Naha city, Japan, based on the mining of frequent congestion patterns that are spatio-temporally continuous, from the long-term observation data of traffic sensors. The results revealed a variety of spread processes of traffic congestion induced by the different traffic conditions, such as weather, time, and day of the week, and confirmed the usefulness of data mining-based approach for understanding the characteristics of traffic congestions occurred in the target area.