Simulation and Physical Experiments Dataset

该数据集包含了为评估AMS-DRL方法在多个追踪者存在情况下指导无人机导航的模拟实验和物理实验。它涵盖了训练过程、推理测试以及实时飞行性能的数据。此外，数据集还包含了多种基准导航策略以供比较，如随机追踪策略、PID追踪策略和手动导航策略。物理实验部分提供了空间几何对成功率影响的分析。规模方面，模拟环境尺寸为5米乘5米乘3米；而物理实验则在现实世界场景中进行。该数据集的任务是多人追踪与规避及导航。

This dataset contains simulated and physical experiments developed to evaluate the AMS-DRL method for guiding unmanned aerial vehicle (UAV) navigation in scenarios with multiple pursuers. It includes data from training processes, inference tests, and real-time flight performance. Additionally, the dataset provides various baseline navigation strategies for comparative evaluation, including random tracking strategy, PID tracking strategy, and manual navigation strategy. The physical experiment section offers an analysis of the impact of spatial geometry on success rate. Regarding scale, the simulated environment has dimensions of 5 m × 5 m × 3 m, while the physical experiments are carried out in real-world settings. The task of this dataset focuses on multi-pursuer tracking and evasion, as well as navigation.