search

CSMRobotics / MATE

MIT License
3 stars 0 forks source link

Redesign Thruster Controller to have dumb controls (priority) and a fancy controller #5

Closed yameatmeyourdead closed 1 year ago

yameatmeyourdead commented 2 years ago
Hermanoid commented 2 years ago

"fancy controls"? I mean, I found the existing stuff to be pretty fancy, what with all the control matrices and attitude control. Does "fancy" mean holding the depth constant automatically? I mean... it would be really nice to try it without that feature, but with all of our neutral buoyancy I think manually managing depth would be super difficult

yameatmeyourdead commented 2 years ago

Fancy controls was referring the the complicated control matrices for attitude and linear velocity control.

It is probably smart to just have automatic depth control as this will be fairly trivial to implement in a PID controller.

Dumb controls would just be direct mapping of joystick to thruster output (rather than the setpoint of a Feedback Controller). This is easily accomplished with some vector math and normalization to get full thrust output. Although this method retains the challenge of mapping 6 DOF to a 4 DOF controller (necessitating translation/rotation modes)

Hermanoid commented 2 years ago

Okay, by this "setpoint of a Feedback Controller" you're alluding to the system you proposed to hold attitude constant while translating linearly and vice versa. In that case, starting with "dumb" controls sounds like a no-brainer.

On Fri, Jul 8, 2022 at 10:35 PM Zac Stanton @.***> wrote:

Fancy controls was referring the the complicated control matrices for attitude and linear velocity control.

It is probably smart to just have automatic depth control as this will be fairly trivial to implement in a PID controller.

Dumb controls would just be direct mapping of joystick to thruster output (rather than the setpoint of a Feedback Controller). This is easily accomplished with some vector math and normalization to get full thrust output. Although this method retains the challenge of mapping 6 DOF to a 4 DOF controller (necessitating translation/rotation modes)

— Reply to this email directly, view it on GitHub https://github.com/Colorado-School-of-Mines-Robotics-Club/MATE_ROV_Programming/issues/5#issuecomment-1179470861, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEIV7VCU2COF3PWEEQWFI2LVTDXQNANCNFSM527K4FGQ . You are receiving this because you were assigned.Message ID: <Colorado-School-of-Mines-Robotics-Club/MATE_ROV_Programming/issues/5/1179470861 @github.com>

yameatmeyourdead commented 2 years ago

Need to run a test to determine accelerometer bias for PID controller stability. Requires works-like testing rig that has mounted BNO to determine integral windup.