SUMO Traffic Microsimulator Data

该数据集来源于使用不同自适应交通信号控制策略的模拟数据，这些策略包括基于学习和非学习的方法，并在动态交通需求场景下生成。数据集涵盖了多种性能指标，如网络旅行时间、交叉口排队长度以及采用不同自适应交通信号控制器的有效性延迟测量（MoE）。该数据集规模宏大，包含了多次模拟（共32次），每次模拟都采用了不同的随机种子。研究任务旨在优化自适应交通信号控制的超参数，并评估控制器的性能。

This dataset is sourced from simulation data generated using diverse adaptive traffic signal control (ATSC) strategies—encompassing both learning-based and non-learning-based methods—under dynamic traffic demand scenarios. It covers a comprehensive set of performance metrics, including network travel time, queue length at signalized intersections, and delay-based Measure of Effectiveness (MoE) for different ATSC controllers. This is a large-scale dataset comprising 32 independent simulation runs, each employing a distinct random seed. The targeted research tasks for this dataset include optimizing the hyperparameters of ATSC systems and evaluating the performance of the traffic signal controllers.