A lane change model under incomplete information game and its control using auction

Abstract

When changing lanes, the driver chooses his or her behavior by predicting the future behavior of surrounding vehicles. In this case, the optimal behavior depends on the extent to which the driver is aware of the behavior of surrounding vehicles. In this study, we formulated the lane change behavior under such incomplete information based on an extensive-form game. Focusing on the difference in cognitive ability between automated and manual vehicles, we attempted to analyze the lane change behavior and traffic conditions under mixed conditions. In addition, we proposed a method to improve traffic conditions in mixed traffic situations through cooperative control of automated vehicles. In order to guarantee the incentive to participate in the cooperative control from the viewpoint of social acceptability, a dynamic pricing method based on auction theory and its theoretical properties are presented.