DYNAMIC REVERSIBLE LANE OPTIMIZATION IN AUTONOMOUS DRIVING ENVIRONMENTS: BALANCING EFFICIENCY AND SAFETY

Reversible lanes are a cost-effective road control method to increase road capacity without significant infrastructure changes. However, their current fixed schedules and low frequency limit their effectiveness in adapting to dynamic traffic patterns. We propose leveraging autonomous driving technologies and cooperative intelligent transportation systems (C-ITS) to enable dynamic operation of reversible lanes. Our study aims to optimize reversible lane location, timing, and installation mode to maximize traffic capacity while minimizing costs, including potential risks and economic expenses. By formulating the problem as a minimum-cost maximum-flow approach for an urban traffic network, we address key decision-making challenges in setting up reversible lanes. We find that choosing the installation mode involves a trade-off between expenses and safety guarantees. We also explore how autonomous driving technology impacts the installation of dynamic reversible lanes. Our results show that dynamic operation with suitable modes significantly improves traffic optimization from both economic and efficiency perspectives. Additionally, autonomous driving reduces risk costs and enhances dynamic adjustment efficiency. We validate the proposed algorithm through comparisons and experiments with real-world traffic data from Xi’an city, demonstrating its practical effectiveness in optimizing reversible lanes.