How could compute joint velocity given end effector twist?

Comments (2)

tkelestemur

I was also looking at this but as I understand from the source code, inverse velocity kinematics is not implemented. It would be great to have to have this feature as well.

2017-10-15T03:11:04+00:00

Patrick Beeson Account Deactivated

changed status to wontfix

TRAC-IK simply solves for joint positions given Cartesian ones. It was meant to be a drop-in replacement for KDL's CartToJnt() functions which takes in Frames and spits out joint values.

You could easily use TRAC-IK (with Minimum Distance solving) to solve for some point in time along a velocity profile of the end effector (say 100ms ahead) and then divide the joint deltas by this time interval to estimate joint velocities. But this is something to implement in the user's code.

2017-12-26T19:14:11+00:00

tkelestemur
I was also looking at this but as I understand from the source code, inverse velocity kinematics is not implemented. It would be great to have to have this feature as well.
- 2017-10-15T03:11:04+00:00
Patrick Beeson Account Deactivated
- changed status to wontfix
TRAC-IK simply solves for joint positions given Cartesian ones. It was meant to be a drop-in replacement for KDL's CartToJnt() functions which takes in Frames and spits out joint values.

You could easily use TRAC-IK (with Minimum Distance solving) to solve for some point in time along a velocity profile of the end effector (say 100ms ahead) and then divide the joint deltas by this time interval to estimate joint velocities. But this is something to implement in the user's code.
- 2017-12-26T19:14:11+00:00
Log in to comment