nunoguedelha
commented
10 years ago Tests have been done on top of master :
- base commit: Merge pull request #77 from bchretien/master (298bd2462d42a98031950e9a98eb0c68f3a9ddca)
- test branch: dbg_check_rnea_issue
== With following section in simple_humanoid.config file (8857c5eca7720b2175b520ac246f42e564c14d37) :
prefer-fixed-axis=true joint=WAIST_P joint=LARM_SHOULDER_P joint=RARM_SHOULDER_P joint=LLEG_HIP_R joint=RLEG_HIP_R
=> RNEA computed torques match the expected values (refer to logs rnea.log, rnea.log.diff) => (H*ddq + C) give wrong torques (refer to logs rnea_from_DynEquation.log, rnea_from_DynEquation.log.diff) These logs are generated by metapod/_build/tests/use_simple_humanoid/test_rnea_simple_humanoid executable.
errors can be significant like:
- WAIST joint (floating) y axis torq error ~ 5 N
- CHEST joint axis torq error ~ 10 N
- most errors are > 1 N
Due to "prefer-fixed-axis=true" option, TransformT sXp template matrices are being instanciated using Spatial::RotationMatrixAboutYTpl
== Without the reffered section (76db02f73af48ae36534fd56199ab8238d9186cb) :
All torques are correct. Torques computed from ID(q,dq,ddq) and H*ddq+C match.
olivier-stasse
If we compute the torques using RNEA, Tau=ID(q,dq,ddq), we get the expected reference torques (rnea.ref). If we use the equation of motion, Tau=H*ddq+C, we get wrong torques. The issue is introduced by the use of optional parameter "prefer-fixed-axis=true" in simple_humanoid.config file . Refer to following comments for more details.