- The paper introduces a two-stage framework that decouples longitudinal and lateral control to enable safe and efficient lane changing.
- It employs iterative target assignment and collision-free path planning to ensure smooth transitions in the lane-changing phase.
- The graph-based MCC scheduling method reduced evacuation times by up to 67.9% and travel delays by 32.6% in simulations.
Cooperation Method of Connected and Automated Vehicles at Unsignalized Intersections
Introduction
The paper "Cooperation Method of Connected and Automated Vehicles at Unsignalized Intersections: Lane Changing and Arrival Scheduling" (2109.14175) addresses the complexities involved in the management of connected and automated vehicles (CAVs) at urban intersections where traditional traffic signals are absent. Recognizing that most existing studies focus on scenarios with prohibited lane changing, the authors present a novel two-stage cooperation framework that accommodates the real-life need for lane changes. This framework decouples the control strategies of CAVs into longitudinal and lateral controls, enhancing both safety and traffic efficiency.
Methodology
The proposed two-stage cooperation framework consists of distinct processes for CAV control and scheduling:
Stage 1 - Lane Changing:
The framework employs a formation control method within a lane changing zone to manage the transition of CAVs to their desired lanes. An iterative solution is applied to address multi-vehicle target assignment and collision-free path planning, circumventing the deadlock problems typical of single vehicle lane changing algorithms. This method aligns well with traffic efficiency goals by ensuring that CAVs smoothly transition to lanes corresponding to their intended trajectories at the intersection.
Stage 2 - Arrival Scheduling:
The second stage focuses on scheduling the arrival of CAVs at the intersection using a graph-based minimum clique cover (MCC) method. This scheduling method optimizes traffic flow by creating a virtual platoon, where CAVs from different lanes can pass through the intersection simultaneously. The MCC method accounts for complex conflict relationships between vehicles, including crossing, diverging, and converging conflicts. This approach reduces computational complexity while optimizing for minimal evacuation time and average travel time delay.
Experimental Results
The effectiveness of the proposed framework is validated through extensive numerical simulations that compare it against constant traffic SPAT. The simulations reveal significant improvements in traffic efficiency across various vehicle numbers and traffic volumes. Specifically, the MCC algorithm was observed to reduce evacuation times by up to 67.9% and average travel time delays by up to 32.6% compared to traditional traffic management methods.
Implications and Future Work
The proposed framework has profound implications for the deployment of CAVs in urban environments, offering a practical solution for managing complex vehicle interactions at intersections with permitted lane changes. Beyond theoretical contributions, the paper suggests that further research could explore the integration of real-time arrival plans initiated at the lane changing zone. Additionally, scenarios involving mixed traffic with human-driven vehicles remain open for exploration, which would broaden the applicability of the proposed system to diverse traffic situations.
Conclusion
This research provides a comprehensive approach for unsignalized intersection management using CAVs through a novel cooperation framework. By facilitating lane changes and optimizing arrival scheduling, the proposed methods enhance both traffic safety and efficiency, demonstrating viability in simulated environments. Future developments could extend these findings to more complex traffic scenarios, potentially revolutionizing urban traffic management systems.