Home
Scholarly Works
Decentralized impedance control
Journal article

Decentralized impedance control

Abstract

We describe an impedance control method following an independent joint control approach. Simple impedance controllers are set up at each joint and coupling among the joints is cancelled using equally simple compensators. It is possible to program any diagonal visco-elastic impedance matrix in Cartesian task coordinates within the limits of what can be achieved at the joint level. The performance depends on the precise tracking of joint torque specifications which may be achieved by minor feedback loops and co-located joint torque sensors. This technique takes advantage of the frequency separation of signals describing the dynamics of most serial robot manipulators to achieve excellent performance and robustness at a low computational cost. In the described method, we should more appropriately speak of control of the impedance rather than of impedance control. Experimental results are reported using a Sarcos redundant manipulator model GRLA (General Robotic Large Arm). The method applies equally well to kinematically non-redundant and redundant manipulators.

Authors

Doyon M; Hayward V; Pelletier M

Journal

IFAC-PapersOnLine, Vol. 27, No. 14, pp. 389–394

Publisher

Elsevier

Publication Date

September 1, 1994

DOI

10.1016/s1474-6670(17)47342-3

ISSN

2405-8963
View published work (Non-McMaster Users)

Contact the Experts team