Multirotor-Gym

该数据集包含了使用具有对抗性重放（AR）的鲁棒强化学习（RL）训练的AR-DDPG策略，以及用于四旋翼无人机位置控制的传统非鲁棒DDPG策略。该数据集还包括了在不同条件下的性能评估，例如无人机质量和外部干扰的变化，以测试训练策略的鲁棒性。规模上，数据集涵盖了多种测试场景，这些场景具有不同的内部和外部参数。任务方面，数据集专注于四旋翼无人机的位置控制及航点导航。

This dataset contains the AR-DDPG policy trained using robust reinforcement learning (RL) with adversarial replay (AR), as well as the conventional non-robust DDPG policy designed for position control of quadrotor unmanned aerial vehicles (UAVs). The dataset also includes performance evaluations under varying conditions, such as changes in UAV mass and external disturbances, to assess the robustness of the trained policies. In terms of scale, the dataset covers diverse test scenarios with distinct internal and external parameters. In terms of task focus, the dataset focuses on position control and waypoint navigation for quadrotor UAVs.