You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I tried to compare using PID and Mellinger for controller and found their yaw direction is opposite.
In cfclient, right yaw commands gives positive deg/s while left yaw gives negative deg/s
with PID controller, cf2 turns right when I commanded right yaw however
with Mellinger controller, cf2 turns left for the same command.
Is this intentional?
I tried to modify some code in the mellinger.c but changing the input does not help solving this issue.
Thanks,
Kyu
The text was updated successfully, but these errors were encountered:
This is not intentional and sounds like a bug to me. The mellinger controller was more heavily tested for operation in a mocap, so this might be related to the slightly separate manual flight mode. I'll take a look.
whoenig
added a commit
to USC-ACTLab/crazyflie-firmware
that referenced
this issue
Jun 23, 2018
The PID and mellinger controllers had different behaviors (inverse
yaw direction) in manual flight mode, see issue bitcraze#335. The root cause
is that the legacy yaw input is actually inverse (with respect to
the CF coordinate system). This change changes the sign accordingly
before processing and fixes some other math issues related
to yaw control.
Tested with manual flight using PID and Mellinger controllers.
I tried to compare using PID and Mellinger for controller and found their yaw direction is opposite.
In cfclient, right yaw commands gives positive deg/s while left yaw gives negative deg/s
with PID controller, cf2 turns right when I commanded right yaw however
with Mellinger controller, cf2 turns left for the same command.
Is this intentional?
I tried to modify some code in the mellinger.c but changing the input does not help solving this issue.
Thanks,
Kyu
The text was updated successfully, but these errors were encountered: