Mixed traffic capacity modeling for connected platoons at unsignalized intersection: A Markov chain-based approach

Emerging technologies have enabled connected autonomous vehicles (CAVs) and connected vehicles (CVs) to achieve platoon driving, which will coexist with regular vehicles (RVs) in mixed traffic. This heterogeneity, characterized by varying levels of automation and connectivity, presents new challenges for traffic capacity modeling. While prior research has primarily focused on highways, capacity modeling of mixed traffic at unsignalized intersections remains underexplored. To address this gap, this study presents an analytical model that assesses the mixed traffic capacity at unsignalized intersections, incorporating stochastic nature of car-following and platooning behaviors among CAVs and CVs upstream of the intersection. A Markov chain-based framework is established to model transitions among different following headway modes, allowing derivations of two key behavioral metrics: car-following probability and platooning probability. These metrics are used to formulate a closed-form capacity model based on average following headways, conflict theory, and queueing principles, reflecting interactions across intersection approaches. To validate the proposed model, numerical experiments are employed to analyze the effects of critical variables on overall capacity, including CAV and CV proportions, maximum platoon length, platooning intensity, and platooning willingness. Results indicate that: (i) The composition of CAVs and CVs plays a critical role in improving mixed traffic efficiency. Under the baseline parameter scenario established in this study, achieving a high throughput requires at least 50% CAVs and 30% CVs at unsignalized intersections. (ii) Increasing platoon size is subject to diminishing marginal returns. Under the experimental conditions of this study, the optimal maximum platoon size is 5 vehicles for CAVs and 3 vehicles for CVs. (iii) Increasing CAVs and CVs’ platooning willingness consistently enhances intersection capacity。These findings provide practical insights into how connected platoons of both CAVs and CVs can enhance mixed traffic capacity at unsignalized intersections.