24.06.03 Progress Meeting

[ssutjipto]

Chair:

@ssutjipto

Attendees:

@Woolfrey
@ssutjipto

---

Announcements

Shoudong is coming to Leeds end of July, Sheila is contact point in Australia

Agenda

Note: Last week's actions become this week's agenda

Item	Responsible	Issue	Outcome
Finish the test script for velocity level joint trajectory tracking	@Woolfrey	#116	Yes, but just need to test a 6DOF robot
Continue on action server	@Woolfrey	Link	Continuing work on this
Type up math for joint level feedback in discussion below	@Woolfrey	#115	Yes, see comments.
Continue working on dynamic control class	@ssutjipto	#117	Still working on it.

Additional Discussion

Topic	Raised By	Issue	Outcome
Make a habit of adding damping torque	@Woolfrey	N/A	N/A
Is the control limit calculation correct	@ssutjipto	N/A	Yes, the equations are correct in the SerialKinematicControl class.
Don't invert matrix directly, if it can be avoided	@Woolfrey	N/A	See comments from this discussion.
Test robustness of the singularity avoidance strategy	@Woolfrey	N/A	In future, will need to test this under different conditions.

Actions

Action	Responsible	Tag
Test 6DOF Cartesian control	@Woolfrey	#122
Continue working on dynamic control class	@ssutjipto	#117
Set default gains as function of control frequency	@Woolfrey @ssutjipto	To be created
Test default gains	@Woolfrey @ssutjipto	To be created

[Woolfrey]

Learn these: https://web.archive.org/web/20240309190945/https://eigen.tuxfamily.org/dox/group__DenseDecompositionBenchmark.html

For example,

Eigen::VectorXd qddot = M.ldlt().solve(tau);

[Woolfrey]

Notes_on_Constrained_Dynamics.pdf