Jam propagation in mixed traffic of autonomous and human-driven vehicles: A random walk-based analysis

Abstract

Traffic jams, characterized by backward-moving waves that disrupt upstream vehicles, are a major concern in transportation that cause congestion and diminish efficiency. This study explores the role of autonomous vehicles (AVs) in mitigating jam propagation in mixed traffic with both AVs and human-driven vehicles (HVs). To capture the complex dynamics of jam propagation in mixed traffic, we develop a novel analytical model grounded in a microscopic perspective to formulate the stochastic jam propagation process and quantify the impact of jam waves in closed-form expressions. Analyses of the closed-form solutions reveal how capacity drops amplify jam waves and identify critical conditions under which AVs can effectively mitigate congestion. Building on these theoretical insights, we use the random walk model to propose enhanced slow-in strategies and validate their effectiveness in mitigating jam propagation through numerical simulations. By modeling the stochastic nature of jam propagation and mitigation, this study contributes to a deeper understanding of AVs’ potential to improve traffic flow, providing a basis for future research on managing mixed traffic systems in the era of AV adoption.