Multi-Agent Reinforcement Learning for Autonomous Vehicle Coordination

The proposed project focuses on the development of a system where autonomous vehicles are modeled as agents, using Multi-Agent Reinforce- ment Learning (MARL) to coordinate their actions. The goal is to opti- mize traffic flow, minimize travel time, and avoid collisions. The system will be evaluated in a simulated environment, measuring key performance metrics such as traffic efficiency and safety. This project will combine the areas of multi-agent systems, reinforcement learning, and autonomous driving to produce a scalable solution for real-world traffic management.