DH, PUMA and Headaches

Help & Support Apps & Software
1

Hello,

I am having a lot of troubles with Denavit Hartenberg parameters (used as a way to describe a kinematics chain made of revolute and prismatic joints) in an IK program. I am trying to compute the forward kinematics for a PUMA type robot arm and while I can get the good position for the first 3 links, the 4th one isn’t working as expected.

In more detail, I use the following DH parameters :

Link# Twist Length Offset Angle

1 0 0 0.5 0
2 -90 0 0 0
3 0 0.5 0.1 0
4 -90 0 0.4 0
5 90 0 0 0
6 -90 0 0 0

If I compute the matrix that give the position/orientation of the fourth link in the frame of the first link, I get the following position :

(0.5, 0.5, 0.5)

The “X” position is accurate, the “Y” isn’t and the “Z” isn’t either. The “Y” should be 0.1 and the “Z” should be 0.1. So it seems that the 0.4 length of the second part of the arm is going on the “wrong axis”, but I have no idea why : not only do the parameters look correct but I have found many different sources giving about the same…

So… This post is a cry for help for anyone comfortable enough with DH to see what is wrong !

Thank you ~

Guillaume

2

This seems like a logical (thinking) error to me… not something necessarily technical. Perhaps if you could more clearly explain the problem without the terminology/shorthand more people would be capable of following your process (without needing direct experience with exactly what it is you’re doing).

atm, what you wrote looks like gibberish to me, but I imagine if you explained how your system “should” work (links to the theories/equations you’re using) and how it “does” work (results from multiple sets of input), I might have been able to offer some suggestions.