In this study, driver behavior is analyzed on the basis of real traffic data measured under varying environmental conditions. In order to achieve safer traffic environments and reduce the number of accidents, it is necessary to clarify driver cognitive processes and determine traveling speeds under a variety of conditions. A driver cognitive model is proposed, in which the driving speed is assumed to be a function of several factors such as overall traveling schedule, speed of the preceding car, road profile, and road surface conditions. An important aspect of the proposed model is that the quality of the information, which is mainly obtained visually, is assumed to be influenced by environmental factors such as visibility. The proposed cognitive model is validated using traffic data measured at a specific location on a northern part of the Tohoku Expressway. The traffic data were collected for one year using a traffic counting system, and weather and road surface conditions were recorded simultaneously. The accumulated data is statistically analyzed. In the preliminary stage of analysis, the relationship between visibility and the distribution of car speeds in the absence of a preceding car is analyzed under different road surface conditions during daytime. During winter, when snowfall occurred frequently, it is shown that car speeds are not affected significantly by visibility when the visibility is over 400 m; rather, they are affected by driver recognition of the road surface, i.e., the different visual contrast induced by the snowfall. This result supports the hypothesis that car speeds are determined by a combination of factors such as visibility and road surface conditions.
ASJC Scopus subject areas
- Control and Systems Engineering