Characteristics and Stability of Car-following Models in Autonomous Vehicles under Mixed Traffic Environments

抄録

This study explores the potential effects of introducing autonomous vehicles in Taiwan's real-world network, utilizing GM (General Motors), IDM (Intelligent Driver Model), and ACC (Adaptive Cruise Control) models as the foundational car-following models to evaluate their influence on mixed traffic conditions. The research uses MATLAB and SUMO to simulate single-lane roads, comparing their performance in modeling traffic flow focusing on the real-world scenario of Zhonghua West Road in Tainan. Simulations demonstrate the GM model's superior stability, while IDM exhibits better stability than ACC and the modified Krauss models. While ACC enhances safety, its impact on traffic efficiency is limited, especially in fully automated scenarios. The research highlights the need for further optimization to balance safety and efficiency in AV-integrated traffic systems. The numerical results provide valuable insights into the dynamics of mixed traffic flow with AVs, paving the way for improved traffic management and optimized AV integration.