Abstract
This study investigates the impact of reverse-variable lane (RVL) length, left-turn traffic arrival rates, green signal timing, and upstream traffic flow on intersection capacity using a Cellular Automata (CA) model. The simulation results reveal a critical threshold for traffic arrival rates: below this value, the implementation of an RVL does not yield significant improvements in left-turn capacity. However, above this threshold, optimized RVL configurations can substantially enhance throughput in congested urban environments.
Addressing Urban Congestion
Urban intersections frequently serve as bottlenecks, particularly for left-turning traffic. In many established city centers, physical constraints prevent the addition of permanent lanes. While traditional methods like extending green lights or creating waiting areas offer some relief, they often reach a point of diminishing returns.
The "reverse-variable lane" (RVL) offers a dynamic alternative. By temporarily repurposing an oncoming exit lane as an additional left-turn lane during specific signal phases, engineers can increase capacity without expanding the physical footprint of the road.
The Concept of Reverse-Variable Lanes
Implementing an RVL requires careful calibration of lane length ($L_r$) and signal timing to ensure safety and efficiency. If the lane is too short, it cannot accommodate enough vehicles; if too long, it may fail to clear before oncoming traffic arrives.
Modeling Traffic with Cellular Automata
To determine the "Goldilocks zone" for RVL implementation, this research utilizes a Cellular Automata (CA) model based on the Nagel-Schreckenberg (NS) evolution rules. This discrete model simulates traffic as particles moving through a grid, accounting for acceleration, deceleration to avoid collisions, human unpredictability (randomization), and position updates. This approach allows for the testing of various scenarios, including different arrival rates and "green times" for left turns.
Simulation Results and Critical Thresholds
The simulation identified a critical traffic arrival rate—approximately 630 veh/h in the tested model—below which an RVL provides negligible benefits. The system only becomes effective once the intersection nears saturation.
As shown in the simulation data, there is an optimal RVL length for any given arrival rate. Beyond this point, increasing the length yields diminishing returns or complicates the clearing process. Interestingly, the study found an inverse relationship between green time and optimal RVL length: longer green lights reduce the pressure on the RVL, allowing for a shorter supplemental lane.
Real-World Application: Yinchuan City
The theoretical findings were validated using a case study of the intersection at Great Wall Road and Phoenix Street in Yinchuan City, China. During evening rush hour, the implementation of a 25-meter RVL increased the south entrance's left-turn capacity by 52.3%. Furthermore, by optimizing the signal timing and using a 35-meter RVL, the researchers cleared the entire left-turn queue in a single cycle while reducing the required green time by six seconds, which could then be reallocated to other congested directions.
Conclusion
Reverse-variable lanes represent a highly effective, low-cost solution for urban traffic congestion, provided they are implemented with precise timing and length. Future work may involve AI-driven management systems to dynamically adjust RVL parameters in real-time based on fluctuating traffic patterns.
Summary updated