In this paper, we present an intelligent intersection control system for multimodal traffic consisting of regular vehicles, connected vehicles, and autonomous vehicles. The system leverages algorithms such as Time to Collision (TTC), Probabilistic Risk Modeling (PRM), Social Value Orientation (SVO), and Deep Q-Network (DQN) to enhance safety and efficiency at intersections. We simulate these algorithms using the SUMO traffic simulator, evaluating their performance under different traffic conditions. Results show that DQN achieves the shortest average crossing time, while TTC maintains a zero-collision rate, demonstrating a tradeoff between safety and efficiency.
Comparative Study of Intersection Management Algorithms for Autonomous Vehicles
01.11.2024
952337 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Mixed intersection network management method considering autonomous vehicles
Europäisches Patentamt | 2024
|A Platoon-Based Intersection Management System for Autonomous Vehicles
British Library Conference Proceedings | 2017
|