Evaluating Stability and Performance in Mixed Traffic: A Theoretical and Co-Simulation Approach

This paper proposes a generalized car-following (CF) model to depict the dynamics of traffic flow that includes human-driven vehicles (HDVs), connected vehicles (CVs) and connected and autonomous vehicles (CAVs). Notably, the model integrates human reaction times, information delays, and status data from multiple preceding vehicles endowed with communication capabilities. Then, by utilizing the perturbation method, the Intelligent Driver Model (IDM) as an example is taken in this CF model to determine the stability condition of the mixed traffic based on CAV penetration rate and their spatial distribution. Finally, comprehensive co-simulation using PreScan and MATLAB/Simulink is developed to explore the impact of varying CAV penetration rates across seven distinct spatial distributions on traffic capacity and dynamic performance. The findings underscore the efficacy of our proposed model in analyzing mixed traffic scenarios comprising HDVs, CVs, and CAVs. Increasing CAV penetration rates can lead to improved stability, capacity, and dynamic performance within mixed traffic environments. Notably, at the CAV penetration rate below 60%, the spatial distribution labeled as CAVs-HDVs-CVs (where CAVs lead the traffic flow, followed by HDVs, then CVs) demonstrates superior dynamic performance, whereas the HDVs-CVs-CAVs configuration (with HDVs leading, followed by CVs, then CAVs) performs worst. However, it's noteworthy that spatial distribution scarcely affects dynamic performance when the CAV penetration rate exceeds 60%.