LLM-POWERED CONSENSUS FOR INTELLIGENT TRANSPORTATION SYSTEM
Keywords: Intelligent Transportation Systems, Distributed Consensus, Large Language Models
TL;DR: This paper introduces an innovative LLM-powered consensus framework for Intelligent Transportation Systems, resulting in improved traffic flow with global optimality and reduced collisions.
Abstract: In the pursuit of Artificial General Intelligence (AGI), the development of Intelligent Transportation Systems (ITS) represents a critical milestone. While current research is largely focused on ego-centric autonomous systems, this approach is limited by the constraints of on-board sensors and local decision-making, leading to potential risks and accidents. Addressing these challenges, this paper introduces an innovative LLM-powered consensus framework for ITS. By leveraging an improved Raft algorithm and incorporating Large Language Models, we facilitate dynamic grouping, inter-vehicle information sharing, and global planning. This decentralized system enhances flexibility, resilience, and the ability to achieve global optimality, resulting in improved traffic flow and reduced collisions. The experimental results validate the effectiveness of our approach in creating a safer and more efficient intelligent transportation ecosystem.
Submission Number: 17
Loading