Prediction of Travel Time Trend on Urban Expressway Using Vehicle Occupancy

Travel time is one of the most important types of information for drivers. However, a travel time can dynamically change according to traffic conditions such as traffic congestion and a traffic accident. It is therefore expected that additional information on travel time information, which shows whether a short future travel time will be increasing or decreasing is needed. In order to predict this travel time trend, the authors first try to define the travel time trend. Then, the authors construct a prediction model using the Support Vector Machine, which judges the travel time trend as “Increase”, “Decrease” or “No change” based on traffic congestion length measured by vehicle occupancies on an urban expressway. As a result of applying the model to real data of Nagoya Expressway in Aichi, Japan, the travel time trend was predicted accurately. Furthermore, these predicted results are verified whether actual travel time of each driver accords with the predicted trend. Then, the authors attempt to predict the difference between the current and the future travel times directly. As a result, the differences can be also predicted accurately as same as the trend predictions.