Understanding altitude hold

ArduCopter
, ,
1

Hi there, can anyone explain me where the D terme and the I term were programmed in this function void AC_PosControl::rate_to_accel_z()
{
const Vector3f& curr_vel = _inav.get_velocity();
float p; // used to capture pid values for logging
// reset last velocity target to current target
if (_flags.reset_rate_to_accel_z) {
_vel_last.z = _vel_target.z;
}

// feed forward desired acceleration calculation
if (_dt > 0.0f) {
if (!_flags.freeze_ff_z) {
_accel_feedforward.z = (_vel_target.z - _vel_last.z) / _dt;
}
else {
// stop the feed forward being calculated during a known discontinuity
_flags.freeze_ff_z = false;
}
}
else {
_accel_feedforward.z = 0.0f;
}

// store this iteration’s velocities for the next iteration
_vel_last.z = _vel_target.z;

// reset velocity error and filter if this controller has just been engaged
if (_flags.reset_rate_to_accel_z) {
// Reset Filter
_vel_error.z = 0;
_vel_error_filter.reset(0);
_flags.reset_rate_to_accel_z = false;
}
else {
// calculate rate error and filter with cut off frequency of 2 Hz
_vel_error.z = _vel_error_filter.apply(_vel_target.z - curr_vel.z, _dt);
}

// calculate p
p = _p_vel_z.kP() * _vel_error.z;

// consolidate and constrain target acceleration
_accel_target.z = _accel_feedforward.z + p;

// set target for accel based throttle controller
accel_to_throttle(_accel_target.z);
}

the schema show there the P and the I and the D whiche it filtred.

Alt-Hold-PID-version-3.0.1.jpg2041×522 46.7 KB

2

I think it is done in the call to accel_to_throttle

https://github.com/ArduPilot/ardupilot/blob/master/libraries/AC_AttitudeControl/AC_PosControl.cpp#L459

The definition in same source file
https://github.com/ArduPilot/ardupilot/blob/master/libraries/AC_AttitudeControl/AC_PosControl.cpp#L464

3

I d’ont think that the pid of the function AC_PosControl::rate_to_accel_z() is in the function void AC_PosControl::accel_to_throttle(float accel_target_z). because if you look the schema there a 3 loops.
first loop contain only P
seconde loop contain PID with the D filtred
the last loop contain PID

i want to know if the code are not the same of the schema or me who doesn’t understand

4

There is no need for PID in the calculation of target acceleration. The PID is used to apply change to throttle to get toward the desired target acceleration

5

i think you don’t understand the concept of PID. try to learn what’s PID.
in the target acceleration. i see only the P term

6

Cheeky monkey. I have been working with PID for 30 years :slight_smile:

http://www.zoomworks.org/oldsite/Winch/BoardElectronics/uber_frame.htm

7

30 years and you did’t understand a simple loop schema. you did’nt see that there is a 3 loop
1 loop only P
2 loop PID
3 loop PID

8

@jazon, please be nice.

9

@jazon
Good luck finding the answer to your question .

Bye