Modify stickbot urdf

[CarlottaSartore]

This PR addresses #15 and changes the links masses as per robotology/icub-models#94, now the robot weight is equal to

idyntree-model-info -m model.urdf --total-mass
[WARNING] SensorElement :: setAttributes : iDynTree does not support sensor of type depth
[WARNING] SensorElement :: setAttributes : iDynTree does not support sensor of type camera
The total mass of model is 52.0102 Kg.

Note that there was the need of recomputing the inertias since the mass changed, I used directly the script inside ergocub gazebo simulator to perform the computation for the changes.
⚠️ the mass of the robot is not exactly equal to the one of icub3 but there are new links (such as the lidar) that increase the mass of the robot.

Since I was already there I have:

• Substitute Nan with zeros for the base_link_fixed_joint; Analyse NaN in stickbot urdf  #8
• Substitute the continuous joint with a revolute one for the r_wrist_yaw and add the limits; r_wrist_yaw is a continuous joint #13
• Removed the masses for the fictitious links l_sole, r_sole, imu_frame, and base_link as it is in iCub3

C.C. @AlexAntn @GrmanRodriguez @Nicogene @lrapetti @DanielePucci

[GrmanRodriguez]

Why was the orientation of this joint changed?

[CarlottaSartore]

Thanks @GrmanRodriguez ! It should not change, I think this was changed by the script. I applied a modification of all the links with a length multiplier of 1 (i.e. no change).
I will investigate why that orientation (and the other origins ) changed! It could be that some parameters of the joint modifier could be wrong (seems to be a flip of direction maybe ? )

[CarlottaSartore]

Fix it in commit 32f38dd

[traversaro]

Are you sure that the orientation changed? Those two set of RPY angles seems to correspond more and less to the same orientation matrix:

mamba create -n idyntreeenv -c conda-forge idyntree
mamba activate idyntreeenv
python
>>> import idyntree.bindings as idt
>>> idt.Rotation.RPY(1.5708,3.1415,0.0).toNumPy()
array([[-9.99999996e-01,  9.26535897e-05, -3.40335638e-10],
       [ 0.00000000e+00, -3.67320510e-06, -1.00000000e+00],
       [-9.26535897e-05, -9.99999996e-01,  3.67320509e-06]])
>>> idt.Rotation.RPY(-1.5707926535897934,9.265358979305728e-05,-3.141592653589793).toNumPy()
array([[-9.99999996e-01,  9.26535897e-05, -3.40335516e-10],
       [-1.22464679e-16, -3.67320510e-06, -1.00000000e+00],
       [-9.26535897e-05, -9.99999996e-01,  3.67320509e-06]])

Note that every single rotation matrix corresponds to multiple euler angles, so if you want to check for equality two set of euler angles you always need to first convert them to rotation matrices, you can't just compare the value of the euler angles.

[GrmanRodriguez]

Were these joint origins wrong?

[CarlottaSartore]

Fix it in 32f38dd

[GrmanRodriguez]

Were these joint origins wrong?

[CarlottaSartore]

Fix it in 32f38dd

[Nicogene]

Merged, thanks!