Inverse dynamics module or extension #129
felixchenier
started this conversation in
2. Development: Ideas, new features, doc, etc.
Replies: 1 comment 1 reply
-
Hey, I was thinking maybe you could use biorbd to run inverse_dynamics. You would have to convert the ktk model into a .bioMod file (easy code) through printing command and run inverse dynamics based on joint angles. Let me know if you are interested in a more detail explanation. It would avoid rewriting and revalidating the complex algorithmic of RNEA. Regards, Pierre |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
One of the most requested features to date is a way to do inverse dynamics calculation. It could be either via a tutorial or an actual module (or extension). There is nothing particularly complex that would prevent it from happening. I also see nothing that could be considered controversial: as long as we make our own assumptions on the body's inertial characteristics, and that the provided method is powerful enough to accept any inertial matrix (e.g., not only a mass and a single radius of gyration), then the rest is purely Newtonian physics.
The difficult part is to create an easy yet complete API to do it. For example, the
kinematics
module has three very simple functions to reconstruct track and reconstruct markers (create_cluster
,expand_cluster
,track_cluster
), that are powerful yet easy to understand. My aim is to get the same simplicity with inverse dynamics.I start this discussion as a notepad for ideas (mine and others), before it turns in a proper feature request. Do not hesitate to put ideas and references.
Beta Was this translation helpful? Give feedback.
All reactions