智能网联环境下 CAv 混行车流集聚策略及分析

A multi-agent system (MAS) based connected autonomous vehicle(CAV) mixed traffic flow aggregation control model (MTF-ACM) was proposed to address the issue of increased conflicts and reduced efficiency caused by the inability of current intersection control methods to adapt to the mixed traffic of connected manual driving vehicles (CHV) and connected autonomous driving vehicles(CAV)． A vehicle-to-vehicle(V2V) based mixed traffic flow aggregation strategy was constructed． To reduce the randomness of mixed traffic flow，a virtual dynamic pre-signal was designed，and the speed of the agglomerated platoon was induced through a spatiotemporal synchronization mechanism. According to the collaborative timing strategy of the main pre-signal，the agglomerated platoon passes through the intersection with the maximum possibility of not stopping． The research results showed that when the market penetration rate (MPR) of CAV is 60%，MTF-ACM achieves the best benefit of capacity，and when the traffic flow is approaching saturation． When the traffic flow approaches saturation，compared to no ACM and CAV-ACM，the average delay time and parking frequency of MTF-ACM decrease by more than 30% and 50%，while fuel consumption and CO₂ emissions decrease by 20.59% and 22.21%，respectively．