Evaluation of Motorway Lane Control Strategies for Mixed Flow of Autonomous and Human-Driven Vehicles

The introduction of automated vehicles (AVs) is commonly expected to improve different aspects of transportation. A long transition period in which AVs will coexist with human-driven vehicles (HVs) is expected until AVs become prevalent. Dedicated lane strategy is considered an effective way to improve road capacity and promote AV use. However, there is a lack of comprehensive research on when and how to implement lane management strategies, and further verification is needed to determine to what extent lane management strategies will affect traffic flow. The dedicated lane strategy will first be applied in highway scenarios, and the merging area is an important zone prone to congestion on highways. There are many impacts of AV on the merging area of highways, but research on the issue that the traffic flow is continually affected after the completion of merging is still lacking. Therefore, this study establishes a lane control strategies framework to investigate the effect on road capacity on the multilane freeway after the merging area. This paper explores the traffic performance of three different lane control strategies with mixed AV/HV traffic flow and investigate when the tested strategies make sense and how sensitive they are to varying AV rates and demands. Specifically, using the open-source microscopic traffic simulation tool SUMO, this study investigates the impacts on traffic performance in terms of throughput, travel time and space mean speed on two-lane motorways at increasing penetration rates of AVs. Moreover, three different lane control strategies (two mixed lanes, one reserved AV lane, and one reserved HV lane) are compared under various demand and AV rates. The simulation results demonstrate that road capacity increases convexly with AV rates. In addition, the results show that the capacity on a one-way two-lane motorway road can be improved with appropriate lane control strategies, especially under high demand and at low to medium AV rates.

The simulation experiments are described in this study, in which a SUMO-based study is designed to evaluate the different capacities for pure HV or AV traffic, and different lane control strategies under different AV rates and traffic demands, together with the results and the traffic performance in terms of changes in capacity, by measuring throughput. We first evaluate the traffic performance of three different lane control strategies with mixed AV/HV traffic flow and investigate when the tested strategies make sense and how sensitive they are to varying AV rates and demands. According to the results, lane strategy can improve traffic capacity. Based on the giving quantized extent of the capacity improvement, the authorities can make decisions on when and how to deploy to dedicated lanes systematically. Lane strategies can significantly improve traffic performance; it should be deployed first on highways, as there is less interference, especially in merging areas, which are prone to traffic congestion. Subsequent testing can be conducted in different road environments to obtain more comprehensive results.