Smoothly adapting viscosity to velocity during comanipulation pointing movements

Author:

Dong Lin,Richer Florian,Roby-Brami Agnes,Morel Guillaume

Abstract

Purpose When a robot comanipulates an object with a human user, damping is a useful function. This is achieved by programming the robot to exhibit a viscous field. For some specific applications, the viscosity is required to change according to the manipulation velocity. A reported method is programming the viscosity varying inversely to the velocity. In this paper, this method is experimentally shown to distort human’s natural motion performance. This paper aims to propose a solution to solve this instability problem. Design/methodology/approach The authors performed a point-to-point targeting movement, where it was observed that the instability results from a sudden reduction of robot’s resistance to motion, which further results from the abrupt viscosity drop when the subject tries to accelerate. Therefore, the authors propose a solution where a first-order linear filter is added to the viscosity coefficient so as to slow down its variation. Findings The experimental results confirm that the proposition is effective, with the ability to stabilize the comanipulated dynamics and to restore the human’s natural behavior. Originality/value This paper concerns applications of comanipulation where the viscosity coefficient is designed to decrease as the velocity increases. An instability problem, which was of vital importance in terms of safety and performance but unreported in the literature, was experimentally studied through human–robot experiments. A solution was proposed by including a secondary dynamics in the variations of the viscosity. Its effectiveness was supported by the practical point-to-point motion experiments.

Publisher

Emerald

Subject

Industrial and Manufacturing Engineering,Control and Systems Engineering

Reference33 articles.

1. Variable impedance control and learning – a review;Frontiers in Robotics and AI,2020

2. The movement of the hand towards a target;Quarterly Journal of Experimental Psychology,1972

3. Quantization of human motions and learning of accurate movements;Biological Cybernetics,1998

4. Motor-cortex-like recurrent neural network and multi-tasks learning for the control of musculoskeletal systems;IEEE Transactions on Cognitive and Developmental Systems,2020

5. Muscle-synergies-based neuromuscular control for motion learning and generalization of a musculoskeletal system;IEEE Transactions on Systems, Man, and Cybernetics: Systems,2020

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3