A Real-Robot Dataset for Assessing Transferability of Learned Dynamics Models

In the context of model-based reinforcement learning and control, a large number of methods for learning system dynamics have been proposed in recent years. The purpose of these learned models is to synthesize new control policies. An important open question is how robust current dynamics-learning methods are to shifts in the data distribution due to changes in the control policy. We present a real-robot dataset which allows to systematically investigate this question. This dataset contains trajectories of a 3 degrees-of-freedom (DOF) robot being controlled by a diverse set of policies. Software to reproduce our benchmark of a few widely-used dynamics-learning methods using the proposed dataset is available in our code repository

在基于模型的强化学习和控制领域，近年来提出了大量学习系统动力学的方法。这些学习模型的目的是合成新的控制策略。一个重要的未解问题是，当前动力学学习方法对于控制策略变化导致的数据分布变化具有何种鲁棒性。我们呈现了一个真实机器人数据集，该数据集允许系统性地探究这一问题。该数据集包含了一个由多种策略控制的3自由度（DOF）机器人的轨迹。使用所提出的数据集，在代码库中可获取用于重现几种广泛使用的动力学学习方法基准测试的软件。