Mixed Autonomy Traffic

Mixed autonomy traffic research focuses on optimizing the interaction between autonomous vehicles (AVs) and human-driven vehicles (HDVs) to improve traffic flow, safety, and efficiency. Current research heavily utilizes reinforcement learning (RL), often employing transformer networks or graph neural networks, to develop control strategies for AVs that account for the unpredictable behavior of HDVs and promote cooperative driving. These efforts leverage both simulated and real-world data, including trajectory data and V2X communication, to train and evaluate algorithms, aiming to create safer and more efficient transportation systems. The ultimate goal is to seamlessly integrate AVs into existing infrastructure, requiring robust and adaptable control systems that can handle diverse traffic conditions and human driving styles.