Micro-macroscopic modeling and fundamental diagram analysis for multilane mixed traffic flow with automated vehicles

The coexistence of Automated Vehicles (AVs) and Human-driven Vehicles (HVs) introduces complex spatiotemporal dynamics into traffic, characterized by heterogeneous car-following behaviors and lane-changing interactions. While microscopic simulations can capture these details, they are computationally prohibitive for large-scale applications. Conversely, macroscopic models often struggle to accurately capture the microscopic heterogeneity and lateral friction effects inherent to multilane operations. To address these issues, this paper proposes a micro-macroscopic modeling framework for multilane mixed traffic flow. High-precision trajectory data from the Waymo Open Motion Dataset and the OpenACC dataset are used to calibrate the Intelligent Driver Model (IDM) parameters for four distinct leader–follower pairs. The results reveal significant behavioral asymmetry across pair types, demonstrating that the microscopic heterogeneity of mixed traffic cannot be adequately represented by a single set of parameters. Building on this calibration, an analytical fundamental diagram of mixed traffic is derived by aggregating the equilibrium states of these microscopic behaviors. Furthermore, a multilane Cell Transmission Model (CTM) is developed by integrating a probabilistic lane-changing mechanism governed by microscopic gap acceptance theory. The proposed model is validated against microscopic simulations in a dual-lane moving bottleneck scenario. The comparative analysis demonstrates that the macroscopic model accurately reproduces key traffic phenomena, including the capacity drop and the propagation of stop-and-go waves. The findings confirm that the proposed framework effectively bridges the micro-macro gap, offering an efficient tool for evaluating the impact of AV penetration rates on freeway capacity and stability.