Quantitative Driver Acceptance Modeling for Merging Car at Highway Junction and Its Application to the Design of Merging Behavior Control

This study models the decision-making characteristics of a driver regarding whether he accepts merging car at highway junction. Then, application modeling to design behavior control is proposed. First, driving on main lane junction observed using simulator, particularly focusing driver's state decision (SOD), which represents acceptance for coming from lane. Second, SOD modeled logistic regression model and prediction performance identified verified. Finally, speed controller designed maximize cars The key idea here minimize entropy by optimizing vehicle. problem quantitatively formulated an addressed applying randomized approach optimization. enables automated vehicle realize considerate Numerical experiments are performed demonstrate usefulness proposed scheme.