Re: Creating D-H table for KUKA youBot

[Peter Corke]

Don't beat yourself up, I think everybody struggles with DH parameters, I certainly still do.  I don't have too much information about this robot, I've just seen some pictures.  Did you make the parameters from first principles or get them from some other source?

How it is "graphed" depends on the joint angles you specify.

To progress this could you:

- send a .m file that builds the robot and plots it

- annotated figure showing what the problem is

Look carefully at the plot, perhaps for a few different sets of joint angles, and check that the joint axes are pointing the correct way in space.  Alternatively, check out the procedure in sec 7.5.2 of my book.

peter

On Sunday, 10 March 2013 12:07:57 UTC+10, Joe.S wrote:

I'm attempting to create a D-H table for the KUKA youBot.
I've been trying to use the Robotics Toolkit to check my work and am coming up with, what I hope is, a problem of misunderstanding on my part...

I've defined my model as
+---+-----------+-----------+-----------+-----------+
| j |     theta |         d |         a |     alpha |
+---+-----------+-----------+-----------+-----------+
|  1|         q1|          0|          0|          0|
|  2|         q2|          1|          0|      1.571|
|  3|         q3|          0|          1|          0|
|  4|         q4|          0|          1|          0|
|  5|         q5|          1|          1|          0|
|  6|         q6|          1|          0|          0|
+---+-----------+-----------+-----------+-----------+
                                                    
grav =    0  base = 1  0  0  0   tool =  1  0  0  0 
             0             0  1  0  0           0  1  0  0 
        9.81              0  0  1  0           0  0  1  0 
                            0  0  0  1           0  0  0  1

When I do this, I expect that the 'link' between joints 5 and 6 would be graphed as "straight out" and that the tool end would have its z-axis pointed "straight out" from the last joint.
The tool does seem to rotate about the z-axis (as it points out from the last joint)... but the link between the 5th and 6th joint is displayed at what looks like an odd angle. (I'll try to post links to pictures of what I expect and a jpeg of the plot.)
Am I defining my parameters incorrectly or is this an artifact of the program? This seems to only occur when I define two joints whose axis of rotation are 90 degrees to each other and have a link of distance "x" between them.
Any help would be appreciated.