Potential field-based modeling and stability analysis of heterogeneous traffic flow

Abstract

Car-following behavior is influenced by multiple factors. Utilizing the potential field theory, previously determined as a comprehensive representation of these influencing factors, this study introduces a unified potential field-based car-following model tailored for mixed traffic flow. Considering the diverse information receptivity of various vehicle types, the model is shaped to encapsulate both Human Driven Vehicles and Connected and Automated Vehicles. Using the characteristic equation-based method, the stability conditions of the model are delineated. Subsequently, a detailed examination of mixed traffic flow stability was conducted across varying Connected and Automated Vehicles penetration rates. Leveraging a dataset of 300 real-world trajectories, the results were benchmarked against classical models such as the Intelligent Driver Model. Notably, our proposed model exhibited an 18.67 % improvement in safety metrics and a 1.36 % improvement in fuel economy over the Intelligent Driver Model when representing Connected and Automated Vehicles. The stability of the mixed traffic flow was found to correlate strongly with vehicular time delay and acceleration parameter weights. Furthermore, augmenting the Connected and Automated Vehicles penetration rate surfaced as a promising strategy to bolster traffic flow stability. Collectively, these findings furnish substantial enhancements to the existing body of knowledge on mixed traffic flow stability, offering pragmatic insights for future research and real-world applications, including traffic management, urban planning, and the development of Intelligent Transport Systems.