Development of human-like automated driving following rules: a comparison between China and Germany

Abstract

ABSTRACT In the evolving autonomous driving field, diverse car-following models and lack of regulations pose challenges. This study establishes human-like car-following rules for different driving styles and conditions across countries. Utilizing Chinese TJRD TS and German High D datasets, Bisecting K-means clustering and ROC curves identified distinct car-following rules. Research showed significant differences in speed and Distance Headway (DHW) among trajectories. Notably, even with the same driving style, speed and instability varied across countries. The study suggests specific Time Headway (THW) settings: in Germany, 1.45 s for conservative and 1.01 s for aggressive driving; in China, 2.25 s for conservative and 1.60 s for aggressive. This research provides insights for tailoring autonomous driving rules to regional characteristics, contributing to the field's development and enhancing understanding of autonomous driving from a human perspective.