0
\$\begingroup\$

I'd like to duple check from units of PID gains for the velocity of car.

This my equations for PI controller

enter image description here

V (m/s), kp (1/sec), ki(1/sec^2), kp' (radians/sec) , ki' (radians/sec^2) and kd (no unit)

I am right?

please help me

Thank in advince

\$\endgroup\$
1
  • \$\begingroup\$ A controller which produces the exact wanted real world driving speed or rotation RPM no matter what's the controlled system would be an invention of the century. Is it possible that you reveal something about the used notation and the structure of the system so that me and others of us who do not have flesh enough between the ears to understand the case fully from 2 equations could get some glue what's the problem? For me your equations look heuristic rules for the wanted driving velocity and rotation speed, not controller's operation equations. \$\endgroup\$
    – user136077
    Commented Apr 15, 2020 at 9:52

1 Answer 1

0
\$\begingroup\$

Attaching units to them is odd to me, but sure, lets give it a crack.

Kp is the error in the velocity so is m/s Ki is (setpoint - kp) * time, so still technically m/s * something, I would leave it unitless kd would be acceleration as it is the rate of change in velocity so m/s/s

Confused how radians enter the discussion unless they are asking you to control the rotational velocity of the wheel which is not usually the same as vehicle velocity (they scale each other based on the radius of any turn the vehicle takes)

If so, kp' would be radians / second kd' would be radians / second / second, And ki' is left with the same above, still radians / second with some odd scaling relation.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.