Kohei WASHIZU

Development of Micro Traffic Simulation Considering Winter Road Conditions

Kazushi SANO

In this study, a traffic simulation was developed to study traffic measures such as snow removal and preventive road closures against the background of long-term vehicle retention caused by vehicles getting stuck due to the recent heavy snowfall.
In developing the traffic simulation, we considered the importance of winter-specific vehicle behavior in order to reproduce winter traffic conditions, and created a model to estimate the desired travel speed of vehicles that takes into account winter road conditions such as road surface conditions, which are thought to have a significant impact on vehicle speed. The speed estimation results were not highly accurate due to the use of road environment data that was not properly discriminated from individual vehicle speeds as the analysis data. As for the explanatory variables, it was found that snow contributed significantly to vehicle behavior because the road surface and the amount of snowfall over a long period of time had a large effect. In addition to the speed estimation model, the vehicle following and lane change models also change model parameters such as distance between vehicles and acceleration/deceleration depending on the road environment. In addition, the model was set to recover speed as the road surface improves to reproduce snow removal operations.
Using the traffic simulation, snow removal operations were analyzed. Since there are no clear criteria for snow removal planning in the current situation, we proposed a method to change the snow removal cycle according to traffic demand and snowfall to obtain a snow removal pattern that is socially and economically efficient and maintains safe and secure road traffic conditions. We also proposed a method to analyze the cost-effectiveness of snow removal operations in terms of benefits and losses. Although the results could not be estimated correctly because the benefits of achieving transportation objectives such as logistics and tourism and the losses due to accidents and stuck vehicles were not taken into account, we believe that this is an effective method.