cfry@watdcsu.waterloo.edu (C.Fry - Inst. Computer Research) (11/23/88)
Inverse Kinematic Functions for Control
of a Redundant Manipulator
by
Charles W. Wampler II
of
Mathematics Department General Motors Research Laboratories War-
ren, MI, USA
Abstract
A manipulator is said to be kinematically redundant if it has a
greater number of joints than the degrees of freedom of its task.
For example, positioning a point on a plane requires 2 degrees of
freedom, so a planar manipulator with 3 or more joints is redun-
dant with respect to that task. The extra degrees of freedom can
be used to change the configuration of the arm while maintaining
the same end-effector position. This can enhance a robot's per-
formance by enabling it to avoid joint limits, obstacle colli-
sions, and kinematic singularities. We have proposed the inverse
kinematic function method to control such robots. The fundamen-
tals of the method will be reviewed and some basic results on
robot workspaces will be presented. A recent implementation of
the method for a 3-joint wrist applied to pointing tasks will be
discussed.
DATE: Wednesday, November 30, 1988
TIME: 3:30 p.m.
PLACE: University of Waterloo, Davis Centre, Room 1302
Everyone is welcome. Refreshments served.