A Short-Term Prediction of Traffic Congestion in a Tourist Area Using ETC2.0 Data

Abstract

This study proposes a short-term prediction method of traffic congestion using ETC2.0 probe data for traffic congestion in a tourist area, which is considered to be highly variable and difficult to predict. Since the demand for tourism in Japan has been increasing steadily, traffic congestion in the vicinity of tourist attractions has become more serious and desired to be improved. In this study, the state-space model was used to predict the travel speed in 500-meter increments and the travel time of the entire section after one hour for the Hakone area of Route 1. In the modeling, after considering various combinations of exogenous variables, the holiday flag and the number of tweets as an indicator to indirectly express the degree of tourism demand are adopted. As a result, it was found that the state-space model was able to predict the speed decrease especially during peak hours.