단속류에서 자율차를 추종하는 비자율차 횡방향 안전성 연구 : 회전교차로를 중심으로

초록

The driving behavior of autonomous vehicles (AV) will affect manually-driven vehicles (MV) in the mixed traffic stream in which AV and MV coexist. This study evaluates the lateral safety of MV following AV at roundabout in the mixed traffic flow and identifies safety vulnerable sections of roundabout. Using SCANeRTM, we built a virtual network based on real roads on a driving simulator, implemented an AV reflecting the characteristics of AV behavior and an MV following the AV. Lateral acceleration, steering velocity, and lane position were selected as indicators to evaluate lateral safety. In addition, safety vulnerable road sections were identified by the proposed sliding window-based evaluation method. As a result of the analysis of lateral acceleration and lane position, the safety vulnerable section was derived from the section before the exit to the roundabout. The proposed methodology was applied to the Paldalmun Rotary located in Suwon city. Regarding the steering angular velocity analysis, the section including the exit from the roundabout was derived as the safety vulnerable section. As a result of deriving the common safety vulnerable section, the section was derived from 11m to 24m upstream from the exit. Using the results of this study, it is possible to identify safety vulnerable sections of roundabouts in MV following AV events. It is expected　to contribute to improving the safety of MV through recommended countermeasures for the derived safety vulnerable sections.

키워드