Li, Dong (Virginia Tech), Wang, Sen (Virginia Tech), Zeng, Haibo (Virginia Tech)

Safe and Efficient Unsignalized Intersection Management with Breadth-First Spanning Tree

Scheduled for presentation during the Regular Session "Autonomous vehicles - intersection management" (FrAT8), Friday, September 27, 2024, 11:30−11:50, Salon 16

2024 IEEE 27th International Conference on Intelligent Transportation Systems (ITSC), September 24- 27, 2024, Edmonton, Canada

This information is tentative and subject to change. Compiled on October 3, 2024

Abstract

Intersections stand as critical nodes in transportation networks, where traffic safety and efficiency are paramount concerns. The advent of Connected and Automated Vehicles (CAVs) holds promise in revolutionizing transportation systems by enhancing efficiency, safety, and congestion mitigation. However, prevailing literature on intersection management adheres to traditional layouts, imposing fixed turning directions on approach lanes, thus limiting the flexibility and controllability of CAVs. In contrast to approaches restricting lane choice, our framework leverages all available exit lanes whenever feasible. The Extended Conflict Directed Graph (ECDG) is proposed to model intersections, integrating vehicle dynamics and intersection geospatial characteristics for improved efficiency. Furthermore, we propose a novel breadth-first-search-based method to solve the problems efficiently. Extensive experiments demonstrate significant improvements in intersection evacuation time, with up to a 15.8% enhancement in traffic efficiency compared to state-of-the-art methods.