Inverse Kinematics with KDL & Newton-Raphson iteration
| Submitted by telvin on Wed, 2011-03-16 18:01 |
Hello!
For Programming of inverse kinematics in my Thesis I want to use KDL. As the test robot I´m using a KUKA KR 125-2. As is known, it is necessary to define the limits of the joints to keep the robot in calculation within the mechanical joint limit. Hence I`m trying to use the Method just like ChainIkSolverPos_NR_JL iksolver1 (chainTest, -3.22885911618951, 3.22885911618951, fksolver1, iksolver1v, 100, 1e-6) but I couldn´t debug my program properly. Is there any concrete example for inverse kinematics programming with joint limits? Even the code example on http://www.orocos.org/kdl/UserManual/kinematic_solvers did not help much. I would be glad if your could give me a feed-back to realize a test.
As the second point I have a question about the calculation of inverse kinematics with the Newton-Raphson iteration. I have read the thesis of Ms. Julia Jesse, which could be found at http://birg.epfl.ch/webdav/site/birg/users/159998/public/report.pdf . According to her calculations the approach of the right positions with respect to the joint limits is not effective enough. Would somebody confirm the results in her thesis at the page 33? Is it possible to improve these values?
I thank you in advance and would appreciate all the helps according to questions above!
Inverse Kinematics with KDL & Newton-Raphson iteration
On Wed, 16 Mar 2011, emre [dot] ates [..] ... wrote:
> Hello!
>
> For Programming of inverse kinematics in my Thesis I want to use KDL. As the test robot I´m using a KUKA KR 125-2. As is known, it is necessary to define the limits of the joints to keep the robot in calculation within the mechanical joint limit. Hence I`m trying to use the Method just like ChainIkSolverPos_NR_JL iksolver1 (chainTest, -3.22885911618951, 3.22885911618951, fksolver1, iksolver1v, 100, 1e-6) but I couldn´t debug my program properly. Is there any concrete example for inverse kinematics programming with joint limits? Even the code example on http://www.orocos.org/kdl/UserManual/kinematic_solvers did not help much. I would be glad if your could give me a feed-back to realize a test.
>
> As the second point I have a question about the calculation of inverse kinematics with the Newton-Raphson iteration. I have read the thesis of Ms. Julia Jesse, which could be found at http://birg.epfl.ch/webdav/site/birg/users/159998/public/report.pdf . According to her calculations the approach of the right positions with respect to the
> joint limits is not effective enough. Would somebody confirm the results in her thesis at the page 33?
> Is it possible to improve these values?
>
> I thank you in advance and would appreciate all the helps according to questions above!
You will have to phrase your questions a lot more precisely and concretely.
With the limited information you currently give, I fear that nobody will be
able to give the "right" answer... For example, you should t say what "not
effective enough" means for you.
This <http://www.catb.org/~esr/faqs/smart-questions.html> might help you
with this difficult process of how to ask the right questions :-)
Herman
Inverse Kinematics with KDL & Newton-Raphson iteration
Hello,
there is a simple difference between both examples. One is without joint limitation and the other one is with joint limitation. The one without joint limitation has better apoximation values. And the question is, if someone would verify this comparision.
Inverse Kinematics with KDL & Newton-Raphson iteration
On Wednesday 16 March 2011 19:01:12 emre [dot] ates [..] ... wrote:
> Hello!
>
> For Programming of inverse kinematics in my Thesis I want to use KDL. As the
> test robot I´m using a KUKA KR 125-2. As is known, it is necessary to
> define the limits of the joints to keep the robot in calculation within the
> mechanical joint limit. Hence I`m trying to use the Method just like
> ChainIkSolverPos_NR_JL iksolver1 (chainTest, -3.22885911618951,
> 3.22885911618951, fksolver1, iksolver1v, 100, 1e-6) but I couldn´t debug my
> program properly.
What do you mean with "I couldn't debug my program properly". The line of code
you provide is wrong syntax, so it probably did not compile. You can check the
right syntax/API on the following page:
http://people.mech.kuleuven.be/~orocos/pub/devel/kdl/latest/api/html/cla...
> Is there any concrete example for inverse kinematics
> programming with joint limits? Even the code example on
> http://www.orocos.org/kdl/UserManual/kinematic_solvers did not help much. I
> would be glad if your could give me a feed-back to realize a test.
The ChainIkSolverPos_NR_JL is very experimental and it does not guarantee to
find a solution. Frankly I would even propose to remove the solver from KDL
until a better solution is provided.
[...]
> I thank you in advance and would appreciate all the helps according to
> questions above!
-- Ruben
--
Orocos-Users mailing list
Orocos-Users [..] ...
http://lists.mech.kuleuven.be/mailman/listinfo/orocos-users