|
Posted by crazygrey on July 25, 2007, 11:09 am
> crazygrey wrote:
> > I'm working on finding the inverse kinematics of the CRS F3
> > manipulator. It is a 6 DOF and I believe there should be a closed
> > algebraic solution for it. I have used the methods suggested in
> > "Introduction to Robotics:Mechanics and Control" by J. Craig and
> > implemented them in MATLAB but I get really complex equations that I
> > don't know how to simplify. I have tried to use "solve" in MATLAB
> > which worked for a 3-DOF before but now it notifies me there is no
> > explicit solution. I have found some methods in the IEEE like the
> > Raghavan and Roth Formulation but it didn't really help. Any help
> > suggestions would be appreciated.
> Matlab is useful for number crunching, not symbolic manipulation.
> For this, try Mathematica; the Robotica package can generate the
> kinematics (and optionally dynamics) directly from the
> Denavit-Hartenberg parameters.
> Grab a copy of Robotica fromhttp://www-cvr.ai.uiuc.edu/~lab/ece470/
> Does this arm use a spherical wrist? If so, the kinematics can easily
> be decoupled and solved as two 3DOF problems.
> Here's another text that may
help:http://www-cvr.ai.uiuc.edu/~seth/index.php?u=spongbook
> Later,
> Daniel
that was very helpful, thanks Daniel. All the joints are revolute. I
will try everything and hopefully it would work.
| 
Yahoo! 
Google 
Windows Live 
del.icio.us 
digg 
Netscape 
XML site map
> manipulator. It is a 6 DOF and I believe there should be a closed
> algebraic solution for it. I have used the methods suggested in
> "Introduction to Robotics:Mechanics and Control" by J. Craig and
> implemented them in MATLAB but I get really complex equations that I
> don't know how to simplify. I have tried to use "solve" in MATLAB
> which worked for a 3-DOF before but now it notifies me there is no
> explicit solution. I have found some methods in the IEEE like the
> Raghavan and Roth Formulation but it didn't really help. Any help
> suggestions would be appreciated.