Bitcraze Forums

Hello.
I'd appreciate it if you could answer my questions.

I saw the code of crazyflie, and I think that crazyflie can perform the comparatively stable flight since the program of PID control is already equipped.
However, even if my copter has not operated it by a controller, the copter doesn’t remain in a fixed point and moves to somewhere.

I thought that the cause of this problem was a gap of the center of gravity etc, so I tried to solve this phenomenon by changing the offset value of Pitch and roll.
But, my copter flies very unconsciously and repeats stopping and moving in the random direction, so I was not able to solve this problem by the above-mentioned method.

If some people experienced condition which resembled this some, please teach me how to solve this problem.
Thank you for your consideration.

Unfortunately, the state-of-the-art is that this is impossible. The onboard controllers measure acceleration and angular velocity, hence getting the position requires to integrate twice. This is suspect to noise and just after a few seconds you get the drift as you described. Calibrating reduces the problem but it won't go away. For longer-term stabilization you need some tracking, for example GPS outdoors, motion capture system indoors, or SLAM based approaches. The later was part of recent research: http://vision.in.tum.de/research/nanocopter.

I am sorry that a reply is overdue.

Thank you for the reply to my question.
I understood that the copter cannot perform perfect stillness, when some tracking was not performed.
Since I was going to read the marker under the copter using the illumination sensor, I try to make the realization of Stable Flight.

Instead of illumination you might be better of with IR photo receptors (and IR LEDs) to avoid issues with ambient light. Another option would be to use a Wii-mote style tracking. If you want to get fancier you can check the various posts in this forum where people used Depth sensors (like Kinect or Asus Xtion Pro) to track the Crazyflie.