SUMO (Simulation of Urban Mobility)

该数据集使用了SUMO交通模拟器作为城市驾驶实验的模拟环境，为涉及车辆导航和变道的学习任务提供了一个强化学习的框架。数据集包含了预定义的路线和用于车辆控制的离散动作空间，这使得各种不同的驾驶场景得以实现，这些场景涵盖了诸如安全和合规等多重目标。该数据集的规模涉及多辆车辆和多种交通情景，其任务是通过多目标强化学习进行城市驾驶模拟。

This dataset employs the SUMO traffic simulator as the simulation environment for urban driving experiments, offering a reinforcement learning framework for learning tasks related to vehicle navigation and lane changing. The dataset contains predefined routes and a discrete action space for vehicle control, which enables the realization of diverse driving scenarios that encompass multiple objectives such as safety and traffic compliance. The scale of this dataset involves multiple vehicles and various traffic scenarios, with the task of conducting urban driving simulation via multi-objective reinforcement learning.