0
TECHNICAL BRIEFS

Torque-Dependent Compliance Control in the Joint Space for Robot-Mediated Motor Therapy

[+] Author and Article Information
Domenico Formica

 Università Campus Bio-Medico, Biomedical Robotics and EMC Laboratory, Via Emilio Longoni 83, 00155 Roma, Italyd.formica@unicampus.it

Loredana Zollo1

 Università Campus Bio-Medico, Biomedical Robotics and EMC Laboratory, Via Emilio Longoni 83, 00155 Roma, Italyl.zollo@unicampus.it

Eugenio Guglielmelli

 Università Campus Bio-Medico, Biomedical Robotics and EMC Laboratory, Via Emilio Longoni 83, 00155 Roma, Italye.guglielmelli@unicampus.it

1

To whom correspondence should be addressed.

J. Dyn. Sys., Meas., Control 128(1), 152-158 (Nov 14, 2005) (7 pages) doi:10.1115/1.2173009 History: Received March 15, 2005; Revised November 14, 2005

This paper is focused on the design of interaction control of robotic machines for rehabilitative motor therapy of the upper limb. The control approach tries to address requirements deriving from the application field and adopts a bioinspired approach for regulating robot behavior in the interaction with the patient. An inner-outer loop control scheme is proposed. In order to tune the level of force and improve robot adaptability in the interaction with the patient, a classical outer force control loop is used. For the inner loop, a novel control law for low-level tuning of robot compliance is introduced, that is borrowed from studies on the biological mechanisms for regulating the elastic properties of the human arm. A dedicated simulation tool, which models the dynamics of an operational robotic machine interacting with a human subject, has been developed. Validation of basic adaptability and safety requirements of the control scheme is carried out in simple tasks, e.g., reaching and contact/noncontact transitions, as well as in simulated situations of typical motor exercises. In particular, the simulation tests demonstrate the adaptive capabilities of the proposed control schemes, e.g., in counterbalancing patient incorrect movements related to the various levels of disability. Moreover, preliminar experimental tests carried out on a real robotic system demonstrated the possibility of using the proposed approach for guaranteeing safe interaction with the patient.

FIGURES IN THIS ARTICLE
<>
Copyright © 2006 by American Society of Mechanical Engineers
Your Session has timed out. Please sign back in to continue.

References

Figures

Grahic Jump Location
Figure 1

Block scheme of the torque-dependent compliance control

Grahic Jump Location
Figure 2

Graphical interface of an operational robotic machine interacting with a human subject: the MIT-MANUS robot arm is drawn in black, the human arm is drawn in gray

Grahic Jump Location
Figure 3

The MIT-MANUS rehabilitation robot (a), and a planar model of the human arm (b)

Grahic Jump Location
Figure 4

Position error in the free space for the torque-dependent compliance control law

Grahic Jump Location
Figure 5

Position error in the free space for the coactivation-based compliance control law

Grahic Jump Location
Figure 6

Interaction force in the constrained motion for the torque-dependent compliance control law. The control gains have the following values: KFP=10−3mN−1, KFI=10−2m(Ns)−1, Fd=1N.

Grahic Jump Location
Figure 7

Interaction force in the constrained motion for the coactivation-based compliance control law

Grahic Jump Location
Figure 8

Pathological (solid line) and normal (dashed line) trajectories (a) and subject trajectories counterbalanced by the robot for Fdx=5N, Ky=100N∕m(b), Fdx=15N, Ky=100N∕m(c), Fdx=5N, Ky=1000N∕m(d), Fdx=15N, Ky=1000N∕m(e) in case of slight disability

Grahic Jump Location
Figure 9

Pathological (solid line) and normal (dashed line) trajectories (a) and subject trajectories counterbalanced by the robot for Fdx=30N, Ky=100N∕m(b), Fdx=45N, Ky=100N∕m(c), Fdx=30N, Ky=1000N∕m(d), Fdx=45N, Ky=1000N∕m(e) in case of severe disability

Grahic Jump Location
Figure 10

Interaction force between the robot arm and the patient in case of severe disability and high control parameters (Fdx=45N,Ky=1000N∕m)

Grahic Jump Location
Figure 11

Norm of the position error in the free space for the torque-dependent compliance control with lower parameters (Rmin=diag{320,60,40,32,16,2,2}Nm∕rad, k=diag{10,10,8,7.2,4,2,2}rad−1, KD=diag{l0,10,6,2,2,0.8,0.8}Nm∕rads−1)

Grahic Jump Location
Figure 12

Norm of the position error in the free space for the torque-dependent compliance control with higher parameters (Rmin=diag{640,120,80,64,32,4,4}Nm∕rad, k=diag{20,20,16,14.4,8,4,4}rad−1, KD=diag{10,10,6,2,2,0.8,0.8}Nm∕rads−1)

Grahic Jump Location
Figure 13

Interaction force in the constrained direction x(Fx)(a), in the unconstrained-direction y(Fy)(b), and in the unconstrained direction y(Fz)(c). The set of force control parameters used is: KFP=10−3mN−1, KFI=10−2m(Ns)−1, Fdx=−5N, Fdy=Fdz=0N.

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In