A centralized relaxation strategy for cooperative lane change in a connected environment

Abstract

This paper proposes a centralized relaxation strategy for cooperative lane changes on the freeway with a connected and automated vehicle (CAV) dedicated lane. Three numerical models are designed in this scenario. Firstly, a lane-changing decision (LCD) model is established considering the speed difference between merging CAV and the target platoon, and the number of following vehicles (FVs) in the platoon. The merging gap is decided by the gap choice model, which is based on the safe distance according to the Gipps model. The relaxation model controls the longitudinal accelerations of all FVs in the platoon to provide proper gaps for merging CAVs. The parameters of these three models are obtained by sensitivity analysis. The results show that safety risks are significantly reduced in the lane-changing (LC) process with centralized relaxation. Benefitting from extending relaxation time, the decrease in amplitude of acceleration alleviates the speed oscillation among the FVs. It also helps to improve traffic efficiency and driving comfort. However, these advantages are weakened when the number of FVs exceeds 5 or the speed difference is over 7 m/s. It is suggested that the single CAV (SV) should merge to the end of the platoon in these conditions from a systematic view.