GEAR_CALIB_T and GEAR_CALIB_P are currently 1.0 in all cases on the Mayall. There may come a date at which robots have variable values set for these. This will affect the internally-tracked POS_T and POS_P angles (which is fine).
However, when running fit_posparams to deduce new values for SCALE_T and SCALE_P, the internally-tracked angles must be corrected for whatever GEAR_CALIB_* values were being used at the time the robot was commanded. Otherwise, the comparison of delta(commanded) vs delta(measured) will come out wrong.
Fix:
1.fitter.py --> Add additional arguments for vectors of GEAR_CALIB_T and GEAR_CALIB_P.
In dynamic mode, pre-adjust the dt_int and dp_int data back to as if GEAR_CALIB_* == 1 at time of move.
GEAR_CALIB_T
andGEAR_CALIB_P
are currently 1.0 in all cases on the Mayall. There may come a date at which robots have variable values set for these. This will affect the internally-trackedPOS_T
andPOS_P
angles (which is fine).However, when running
fit_posparams
to deduce new values forSCALE_T
andSCALE_P
, the internally-tracked angles must be corrected for whateverGEAR_CALIB_*
values were being used at the time the robot was commanded. Otherwise, the comparison of delta(commanded) vs delta(measured) will come out wrong.Fix: 1.
fitter.py
--> Add additional arguments for vectors ofGEAR_CALIB_T
andGEAR_CALIB_P
.dt_int
anddp_int
data back to as ifGEAR_CALIB_*
== 1 at time of move.