Gravitational component of dyamic model about rbdl HOT 4 CLOSED

Hi there,
I hope you understand that we cannot help you solving the problems related to your master thesis.
However, we can maybe give you some directions. Have you looked at the documentation for said algorithms (InverseDynamics/NonlinearEffects) here?
RNEA solves the following problem given q, qdot, qddot it computes generalized joint forces, i.e.,
RNEA(q, qdot, qddot) = M(q)qddot + n(q, qdot) = tau
For example, NonlinearEffects (tailored RNEA for qddot = 0), computes the term n(q, qdot). It accounts for the Coriolis and centrifugal forces, gravity, and any other forces acting on the system other than those in tau. If you now play around with a simple example you should be able to find out how to compute the gravity term g(q) using NonlinearEffects.
Maybe @martinfelis or @mjhmilla can also comment on it.
Best
Manuel

from rbdl.

Ok, I already understood how the algorithm works. If I only give position values ​​the value of tau is equal to the gravitational component, don't?
And I should clarify that I think I expressed myself wrong, this is for a task of a subject of the master, not for the thesis 😅

from rbdl.

I think you're still missing something:

N(q,qdot): contains gravity, centrifugal, and Coriolis forces
tau: contains only the generalized forces that you are applying to the model, most commonly from from motors/muscles.

If you read Featherstone's text, or some of his papers, the grouping of all of the 'bias' terms into N(q,qdot) should be clear. Not every formulation of the equations of motion groups the bias forces together in this way which I expect is causing your confusion.

from rbdl.

I think that the question is answered.
Thanks for everyone's feedback.
Good luck with your masters.

from rbdl.