On-ramp mixed traffic flow merging model with connected and autonomous vehicles

The impacts of merging on mixed traffic flow in the highway on-ramp zone are numerically investigated in this paper. An on-ramp mixed traffic flow merging model is proposed. Our model focuses on the on-ramp merging zone and considers a mixed traffic flow comprising of connected and automated vehicles (CAVs) and human-driven vehicles (HDVs). Firstly, we establish the Velocity-Modified Comfortable Driving (V-MCD) model for HDVs, which incorporates a velocity estimation module into the base Modified Comfortable Driving (MCD) model. The Cooperative Adaptive Cruise Control (CACC) model developed by PATH Labs is used for CAVs. A cooperative merging model has been developed to capture the interaction behavior of vehicles in various driving scenarios. Merging rules are established based on vehicle type and merging order. The merging rules for HDV-HDV are based on safety gap. Meanwhile, those for CAV-HDV are based on velocity difference, and for CAV-CAV are based on first-come-first-served. We conduct numerical simulation experiments to derive the fundamental diagram of the mixed flow and analyze the interaction between vehicles on the on-ramp and mainline. The simulation results indicate that the V-MCD model can effectively reflect on the hysteresis effect caused by on-ramp merging vehicles in real traffic scenarios. The increase in CAV penetration improves road traffic efficiency in terms of both traffic velocity and capacity. However, as the penetration of CAVs continues to rise, their positive impact will diminish. The sensitivity analysis of the hysteresis effect at both micro and macro levels can enhance our understanding of the impacts on traffic flow. As the hysteresis effect increases, merging has a greater impact on mainline upstream traffic, leading to reduce headway and more erratic velocity changes.