arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2109.04743cs.RO
收藏

一种适用于单边约束下任意任务维度的人-机器人物理交互层次框架

A Suitable Hierarchical Framework with Arbitrary Task Dimensions under Unilateral Constraints for physical Human Robot Interaction

KUKA Germany GmbH(库卡德国有限公司);Leibniz University of Hannover(汉诺威莱布尼茨大学)

Juan David Munoz-Osorio, Felix Allmendinger

摘要:

在过去的几年中,已经提出了一些层次框架来处理高度冗余的机器人系统。其中一些系统预计将执行多项任务,并与环境进行物理交互。然而,没有一个建议的框架是能够管理多个任务与任意的任务尺寸,同时尊重单边约束的位置,速度,加速度和力的水平,并在同一时间,直观地反应到外部力量。这项工作提出了一个框架,解决这个问题。该框架进行了测试,在模拟和一个真正的机器人。冗余协作工业机器人(KUKA LBR iiwa)的实验证明了该框架的优势相比,国家的最先进的方法。该框架直观地对外部力作出反应,并且能够限制关节位置、速度、加速度和力。

英文摘要:

In the last years, several hierarchical frameworks have been proposed to deal with highly-redundant robotic systems. Some of that systems are expected to perform multiple tasks and physically to interact with the environment. However, none of the proposed frameworks is able to manage multiple tasks with arbitrary task dimensions, while respecting unilateral constraints at position, velocity, acceleration and force level, and at the same time, to react intuitively to external forces. This work proposes a framework that addresses this problem. The framework is tested in simulation and on a real robot. The experiments on the redundant collaborative industrial robot (KUKA LBR iiwa) demonstrate the advantage of the framework compared to state-of-the-art approaches. The framework reacts intuitively to external forces and is able to limit joint positions, velocities, accelerations and forces.

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。

↑