01-11-2024, 03:05 PM (This post was last modified: 01-11-2024, 04:36 PM by Albert.)

Hi,
I have this problem, when calculating Forward Kinematics in Matlab, the calculated matrix is not the same as the desired matrix in RoboDK. I tried three different solvers including RoboDK one and the problem occur at all three - deviation along x and y axis. Do you know why this happens and how to solve this problem? I am attaching pdf file for more info.

01-11-2024, 03:21 PM (This post was last modified: 01-11-2024, 03:22 PM by Albert.)

The output of RoboDK's SolveFK (Forward Kinematics solution) and the input of SolveIK (Inverse Kinematics solution) are in mm.

It looks like the Peter Corke kinematics library you are using uses meters instead. So you simply need to multiply the position cells of the pose by 1000 (the last column).

I am aware of the different units, but as you can see from the pictures, even using Robodk's SolveFK with joints in deg and Tool offset in mm the values are still not ok. X and Y of the T3 should be around 89 and -526 mm but are instead 92 and -537, making Y value go off by 11 mm, which can be quite a lot in some cases. I put the Corke's and IKFast calculation in T1 and T2 just for reference, because they also deviate from values in RoboDK (T2 example: desired y: -526,922, actual Y in m: -0,5380, actual Y in mm: -538).

Ok, I have tested Forward Kinematics in empty cell with only the robot without the tool and the values are correct in all three cases. That means that deviation must happen somewhere at my end on the tool or coordinate systems.