Add inertial parameters regressor method to KinDynComputations #480
Conversation
|
With this commit, the |
traversaro
changed the title
doc/dcTutorialCpp.md
Outdated
| // Check that ok is actually true | ||
|
|
||
| KinDynComputations kinDynComp; | ||
| kinDynComp.loadRobotModel("model.urdf"); |
There was a problem hiding this comment.
This should not work!
kinDynComp.loadRobotModel(mdlLoader.model());
There was a problem hiding this comment.
I have toooo many failed attempts to get good documentation around, I should just thrown them our. :D
Anyway, nice catch!
| /** | ||
| * Compute the getNrOfDOFS()+6 vector of generalized external forces. | ||
| * | ||
| * The semantics of baseAcc, the base part of baseForceAndJointTorques |
There was a problem hiding this comment.
IMHO using "variable name style", i.e. baseAcc and baseForceAndJointTorques in a function that does not accept these variables is a bit confusing.
Same for the inverseDynamicsInertialParametersRegressor method.
There was a problem hiding this comment.
Yes, that's a left over, thanks for catching that.
|
Pushed the docs of the new method, and some general improvements of the docs of KinDynComputations. The documentation will probably need many more improvements, but this is a first step. Example:
|
|
I think I need to stop in trying to fix all the docs otherwise I will do that for one month. : ) |
In particular, add the inverseDynamicsInertialParametersRegressor that computes the (6+nrOfDofs) \times (10*nrOfLinks) matrix that maps the inertial parameters to the part of inverse dynamics that depends on inertial parameters, and the generalizedExternalForces that computes the contribution of external forces to inverse dynamics (useful because those are not included in the torques computed by the regressor).
Cleanup also documentation for KinDynComputations, to simplify the process of adding docs for the new method.
traversaro
force-pushed
the
addRegressor
branch
from
eaac853 to
122beec
Compare
In particular, add the inverseDynamicsInertialParametersRegressor that
computes the (6+nrOfDofs) \times (10*nrOfLinks) matrix that maps the inertial parameters
to the part of inverse dynamics that depends on inertial parameters,
and the generalizedExternalForces that computes the contribution of external
forces to inverse dynamics (useful because those are not included in the
torques computed by the regressor).
I would still like to add some more docs, but nevertheless the code and the tests should be complete.