TUM-Aries-Lab / exosuit-python / 35599196164
88%

DEFAULT BRANCH: main
Ran
Jobs 1
Files 10
Run time 1min
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst
coverage: 88.276% (+0.1%) from 88.161%
35599196164

push

github

NGierden 
Tune the position-loop gain for the motor this suit actually has

kp came from motor_control.py, where KP_MAIN is 16.0. That rig is an AK60-6:
its decode constants are P +/-12.5, V +/-45, T +/-15, an exact match for
AK60_6_V1_1_MIT_LIMITS, and its header cites the AK60-6 manual. This suit runs
an AK80-6 -- 6.0 N*m rated against 3.0, 12.0 peak against 9.0, 21 pole pairs
against 14, and a much heavier rotor.

The gain sets the velocity trajectory the motor is asked to follow, and the
torque to follow it is J * alpha, so the same number asks a bigger rotor for
proportionally more. Measured against the rig's no-classif ID_07 run, torque
while the motor was moving was 3.48 / 3.78 N*m here against 1.78 / 1.38 there,
while torque at rest matched at 0.58 / 0.86 against 0.50 / 0.41 -- which is
what places it in the acceleration term rather than in tension. The motor
reversed direction 33.9 / 20.7 times a second against the rig's 15.4 / 14.3:
it lags, overshoots and comes back, which the wearer felt as reacting to every
small movement.

16 * (1.6 / 3.6) is about 7, so matching the rig's moving torque lands on the
Simulink model's original 8.0, which is also hip-controller's PIDConfig
default. The comment that called that the stale number was wrong: it is the
one that fits this motor.

Both values are named in the docstring, and they bracket the useful range --
raise it back towards 16 if the assist feels weak or tracking goes loose.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

768 of 870 relevant lines covered (88.28%)

0.88 hits per line

Jobs
ID Job ID Ran Files Coverage
1 35599196164.1 10
88.28
GitHub Action Run
Source Files on build 35599196164