0
$\begingroup$

I have a Simulink model that has a standard Simulink block that determines the Jacobian for a given Rigid Body Tree model. The Rigid Body Tree model is a built in model of the Franka Emika Panda robot and is loaded into the workspace using: panda_matlab = loadrobot("frankaEmikaPanda").

enter image description here

The input to the block consists of the robots configuration (config), which is determined by the joint angles $q$. In this case all the joint angles are $0$. This means that the robot is in an upright position and joints 3 and 5 depend linearly on joint 1 Which means that columns 1, 3 and 5 of the Jacobian should be identical. This is indeed the case, However the Jacobian also has one row of zeros. I do not understand where this row of zeros come from.

enter image description here

I also used my own, simplified Panda model with fewer links. In the same upright position with, the same linearly dependent links. The jacobian matrix of this model is shown below. This also has a row of zeros in the Jacobian. enter image description here

What is causing the jacobian to have this row of zeros?

Thanks in advance.

Improve this question
$\endgroup$
4
  • 1
    $\begingroup$ between which frames are you computing the jacobian? $\endgroup$
    – FirefoxMetzger
    Oct 8 at 14:13
  • $\begingroup$ Base frame to end-effector. For a Panda robot in upright position. $\endgroup$
    – user463102
    Oct 11 at 9:05
  • 1
    $\begingroup$ Then the two columns of 0 should be panda's finger joints (which you don't have in your own simplified model). They will not contribute to a pose change of the end-effector frame and will hence be a column (not row) of 0s. $\endgroup$
    – FirefoxMetzger
    Oct 11 at 16:48
  • $\begingroup$ I understand it better now. Thanks a lot. $\endgroup$
    – user463102
    Oct 12 at 12:11

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.