Page 1 of 1

Yaw angle behavior

Posted: Fri Aug 03, 2018 2:34 pm
by rafael.mjas
Hello,

Recently, we've bought the Crazyflie 2.0 quadcopter and we have a few questions about the yaw angle controller.

Since the Yaw PID controller is originally disabled on the firmware, because of the fact the its measurement comes only from the gyro, what makes it less accuracy, the question is: How does the quadrotor is able to return to its original yaw angle when, in hover, it's submitted to an impulse momentun (on the yaw direction)? How the controller performed this behaviour since it comes originally disabled?

Thank you.

Re: Yaw angle behavior

Posted: Mon Oct 08, 2018 9:47 am
by arnaud
Crazyflie always has a YAW controller enabled so I am not completely understanding the question but I will try to explain a bit how things are supposed to work, do not hesitate to follow up if I am not answering the question.

Technically, by default, the PID yaw control is setup in such a way that absolute yaw angle control is always enabled, velocity control on the yaw is simulated by 'moving' an internal absolute yaw setpoint around. This was implemented to minimize yaw drift.

If you are not using any absolute positioning system (ie. LPS. Mocap, ...), the YAW is estimated by accumulating gyroscope measurement. If you are using an absolute positioning system, the Kalman filter will also estimate Yaw using the flight dynamic.