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 1807.09904cs.ROcs.LGcs.SYeess.SY
收藏

一种数据高效且精确可控的推动作方法

A Data-Efficient Approach to Precise and Controlled Pushing

Department of Mechanical Engineering — Massachusetts Institute of Technology(机械工程系——麻省理工学院)

Maria Bauza, Francois R. Hogan, Alberto Rodriguez

AI总结:

本文提出了一种数据高效的方法,通过学习动态模型来控制复杂机械系统,仅需10个数据点即可完成复杂的推动作轨迹。

摘要:

几十年来,控制理论的研究表明,简单的控制器在获得及时反馈的情况下,能够控制复杂的系统。推动作是复杂机械系统的一个例子,由于摩擦系数和压力分布等未知系统参数,难以准确建模。本文探讨了控制而非建模所需的数据复杂性。结果表明,一种基于模型的控制方法,其中动态模型从数据中学习,能够使用极少量的训练数据(10个数据点)完成复杂的推动作轨迹。推动作的动态特性通过高斯过程(GP)建模,并在一种模型预测控制方法中利用,该方法线性化GP并施加执行器和任务约束,以完成平面操作任务。

英文摘要:

Decades of research in control theory have shown that simple controllers, when provided with timely feedback, can control complex systems. Pushing is an example of a complex mechanical system that is difficult to model accurately due to unknown system parameters such as coefficients of friction and pressure distributions. In this paper, we explore the data-complexity required for controlling, rather than modeling, such a system. Results show that a model-based control approach, where the dynamical model is learned from data, is capable of performing complex pushing trajectories with a minimal amount of training data (10 data points). The dynamics of pushing interactions are modeled using a Gaussian process (GP) and are leveraged within a model predictive control approach that linearizes the GP and imposes actuator and task constraints for a planar manipulation task.

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。

↑