arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1703.00472cs.ROcs.LG
收藏

旋转任务的强化学习

Reinforcement Learning for Pivoting Task

KTH Royal Institute of Technology(瑞典皇家理工学院)

Rika Antonova, Silvia Cruciani, Christian Smith, Danica Kragic

摘要:

在这项工作中,我们提出了一种方法来学习一个强大的政策,解决旋转任务。最近,一些无模型连续控制算法被证明可以在没有任务动态的先验知识的情况下学习成功的策略。然而,获得成功的策略需要数千到数百万的训练集,限制了这些方法对实际硬件的适用性。我们开发了一个训练程序,允许我们使用一个简单的自定义模拟器来学习对模拟与机器人不匹配的鲁棒策略。在我们的实验中,我们证明了在模拟器中学习的策略能够将对象旋转到真实机器人上所需的目标角度。我们还展示了对具有不同惯性,形状,质量和摩擦特性的对象的泛化,而不是在训练过程中使用的对象。这一结果是使无模型强化学习可用于解决机器人任务的一个步骤,通过在模拟器中进行预训练,只提供与现实世界动态的不精确匹配。

英文摘要:

In this work we propose an approach to learn a robust policy for solving the pivoting task. Recently, several model-free continuous control algorithms were shown to learn successful policies without prior knowledge of the dynamics of the task. However, obtaining successful policies required thousands to millions of training episodes, limiting the applicability of these approaches to real hardware. We developed a training procedure that allows us to use a simple custom simulator to learn policies robust to the mismatch of simulation vs robot. In our experiments, we demonstrate that the policy learned in the simulator is able to pivot the object to the desired target angle on the real robot. We also show generalization to an object with different inertia, shape, mass and friction properties than those used during training. This result is a step towards making model-free reinforcement learning available for solving robotics tasks via pre-training in simulators that offer only an imprecise match to the real-world dynamics.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑