Modeling of Expressway Merging Behavior for Autonomous Vehicle Control

Abstract

The purpose of this research was to study the process of formation of collaborative merging behavior in an expressway-merging-area, and to propose merging schemes for autonomous vehicles based on this process. Vehicle positions and speeds were analyzed with high accuracy using digital image processing to videos of vehicles in an expressway-merging-area taken from an elevated position. Three merging patterns were identified as a result. The results suggest that merging patterns are decided collaboratively, with drivers taking into consideration whether the priority is for direct merging or for yielding when the merging vehicle is at the tip of the hard nose.