Heterogeneous traffic flow theory fundamental diagram model considering vehicle queuing characteristics and lane changing behavior

Lin Hou, 宋贵玉, Yulong Pei, Weizhi Dong, Heyao Gao, Xiujuan Tian

The emerging heterogeneous traffic environment comprising human-driven vehicles (HDVs) and connected autonomous vehicles (CAVs) necessitates theoretical tools to explicitly capture queueing and lateral interactions. This study develops a Heterogeneous Fundamental Diagram (FD) based on the Multi-Lane Multi-Vehicle Interaction Sensitivity Function (MLMVISF) model, incorporating vehicle following ratios derived from Markov chains and embedding fleet strength and maximum fleet length constraints. Different automation levels and lane-changing behaviors of HDVs are incorporated through safety-distance-based rules that feed back into longitudinal dynamics. Numerical experiments spanning a wide range of CAV penetration rates, platoon density, maximum platoon length, and lane-change intervals reveal that increasing CAV platoon density shifts the fundamental diagram upward and delays congestion onset. Extending maximum platoon length from 4 to 10 vehicles only marginally increases maximum flow from 4306 to 4507 vehicles per hour, revealing an effective upper bound on platooning benefits. Reducing lane change spacing is critical for maintaining capacity. The resulting baseline provides a coherent analytical-numerical framework for evaluating cooperative control strategies in heterogeneous traffic flows.