Variable Speed Limit Control Method on Freeway Merging Area Under Car-Truck Mixed Heterogeneous Traffic Condition

Abstract

To address traffic congestion and safety challenges in freeway merging areas under car-truck mixed heterogeneous traffic conditions, this paper proposes a variable speed limit (VSL) control method. Considering a one-way, two-lane freeway merging area as the research scenario, a lane-level cell transmission model (CTM) is developed to predict traffic state under car-truck mixed heterogeneous traffic conditions, incorporating four vehicle types: human-driven cars (HCs), human-driven trucks (HTs), connected autonomous cars (CACs), and connected autonomous trucks (CATs). A collision risk prediction model for the merging area is constructed based on simulation data. A VSL control function is formulated and integrated into a model predictive control (MPC) framework to optimize traffic efficiency and safety, thereby establishing a VSL control method for freeway merging areas. The effectiveness of the proposed method is validated through simulations conducted with Python and SUMO. Simulation results demonstrate that, compared to the no VSL control condition, the VSL method significantly reduces both total travel time (TTT) and collision risk (TCR) at the macroscopic level. Specifically, when the truck mixing rate is 0.2, and the CAV penetration rate reaches 0.5, the TTT decreases by 8.24%, and the TCR decreases by 5.93%. At the microscopic level, compared to the no VSL control condition, the average speed increases, whereas travel delay, time-exposed TTC (TET), and time-integrated TTC (TIT) decrease, further confirming the effectiveness of the proposed method. The proposed research framework provides theoretical insights and technical support for future freeway traffic management and control under car-truck mixed heterogeneous traffic conditions.

Keywords: Intelligent transportation system, variable speed limit control, merging area, car-truck mixed heterogeneous traffic flow, model predictive control