2024 Volume 80 Issue 20 Article ID: 24-20111
In recent years, the logistics industry has been promoting more efficient transportation using large vehicles, and the number of applications for traffic permits for special vehicles has been increasing. In order to reduce the burden on carriers and road administrators and to speed up the screening process, there is a need to automate the screening process. In this study, a database of intersection characteristics and information on whether special vehicles are allowed to turn or not was compiled, and a model for automatic determination of whether a vehicle is allowed to turn or not was constructed. The results of a basic tabulation of intersection parameters manually read from Google Map, etc., showed a clear relationship between the intersection area, road width, turning angle, and other parameters and whether or not a vehicle can turn or not. Next, a model for determining whether or not to proceed was developed. After comparing several methods, the neural network model showed the highest accuracy, exceeding 0.7 for vehicle class 0.
