Reliability-Based Equilibrium Model Considering Promotive Impacts of Connected and Autonomous Vehicles on Traffic Flow Stability in a Mixed Traffic Network

Existing studies on traffic flow stability primarily focused on local stability, with little attention given to its extension to the network level, known as network stability. In this paper, a reliability-based equilibrium model in a mixed traffic network including human-driven vehicles and connected and autonomous vehicles is developed to analyze the impact of connected and autonomous vehicles on traffic flow stability. The basic characteristics of the model are first examined on a small network, demonstrating the non-uniqueness of the link flow in the user equilibrium pattern. Then, the model is extended to the case of a general network with Variational Inequality (VI) equations. In addition, a two-level optimization strategy is developed by incorporating the pricing and quantity control strategies to the reliability model. Numerical examples are conducted based on Sioux Falls networks to examine the performance of the proposed models.