The study about capacity drops of CAVs system under a heuristic rules-based algorithm at minor-level merging on multi-lane highways

Merging bottleneck, as a typical urban traffic scenario, has gained a growing attention. According to our previous research about the merging of CAVs system on two-lane highways, a heuristic rules-based algorithm using virtual vehicle is applied and the outflow rate finally remains the maximum capacity. As a continuation, this paper further investigates the traffic evolution on multi-lane highways. Taken three-lane highways as a typical scenario, vehicles will experience major-level merging (main lane) and the minor-level merging (two branch lanes). Different with two-lane highways, capacity drop occurs during minor-level merging and acts as a redistribution of flow rate when merging to main lane. When inflow on main lane increases, the outflow after minor-level merging is dynamically balanced to remain the total outflow rate equal to the maximum capacity. In simulation, some traffic performances are employed to investigate the performance indicators of traffic system, such as travel delay, fuel consumption and monetary cost. The results demonstrate that capacity drop during minor-level merging comes from the allocation of road rights in proposed control algorithm, which improves the traffic performance of CAVs on main lane by sacrificing the driving of CAVs on two branch lanes.