Quantcast

TranslationDirection5D with Custom 5dof

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

TranslationDirection5D with Custom 5dof

Humpelstilzchen
Hi,

I'm currently trying Openrave (git c0d38fc with sympy 0.7.1) on my Custom 5dof. Unfortunately the ik seem to fail a lot, even with positions that should be physical reachable, the call to

openrave.py --database inversekinematics --robot=arm.xml --iktype=translationdirection5d --iktests=100 --show

ends up with a lot of these errors:
[ikfastsolver.cpp:1625 _ValidateSolutionAll] ignoring bad ik for arm:gripper dist=0.04004096660314603, param=[1442840583 -0.1987926843695766 -0.9722928033231021 -0.1229966391705129 0.2652216005701876 -0.162347375041132 -0.1226124960969151 ], sol=[3.69087609310095, 3.484419979695062, 1.023237518019122, 4.90853500302848, 0.3133917207171269, ]
openravepy.databases.inversekinematics: show, found 0 solutions

which seem to indicate that the ik doesn't seem to work. In the case that someone has an idea what might me wrong I have uploaded my dae[1].

Btw: the wiki seems to be down. I only get a php Fatal error.

Greetings,
Erik

[1] http://vontaene.de/ros/openrave/arm.zip


------------------------------------------------------------------------------

_______________________________________________
Openrave-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/openrave-users

signature.asc (188 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: TranslationDirection5D with Custom 5dof

Humpelstilzchen
ok, got it working by making the following changes:

- Doing all rotations about the Z-axis (as in the Denavit-Hartenberg convention)
This change got rid of the false results ("ignoring bad ik for" error).

- Placing the end effector at an offset from the last joint.
In my robot description I have placed the last link that my end effector attaches to directly on the last joint.
This is of course not correct, moving it away from the the last joint gave me valid ik solutions.

Also if the manipulator direction is not configured OpenRave seems to default to the Z-axis.
Hope this helps someone.

greetings,
Erik


------------------------------------------------------------------------------

_______________________________________________
Openrave-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/openrave-users

signature.asc (188 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: TranslationDirection5D with Custom 5dof

lee520
In reply to this post by Humpelstilzchen
<P>Less shield, Understanding that guides cash flow thurs, Is being successful with this <STRONG>Exercices Trx</STRONG> place. Specials until this year are motivated to be a lot greater than double what we were looking at near 2012. Vehicle gives were able to keep profit margins the <STRONG>Trx Training</STRONG> highest even <STRONG>Exercice Trx</STRONG> if even then using purchases <STRONG>Trx Exercises</STRONG> in the commercial. </P>
<P></P>
Loading...