Bitcraze Forums

Hi crazyflie team, i'm working on a research project utilizing the Crazyflie 2.0 quadcopter, z-ranger, and Loco Positioning System (8 Nodes). I've been trying to obtain accurate position readings utilizing a Python script similar to the one called "AutonomousSequence". Is there a way to use the z-ranger only to calculate the z cordinate and x,y with the lps?

Sometimes, I believe when the sensors (LPS and z-ranger have too far results on z), the crazyflie goes crazy and can no longer maintain the hover position or the assigned trajectory.

Thanks in advance.

The current implementation of the kalman filter is supposed to use the z-ranger reading when the Crazyflie is close to the floor and gradually start using more LPS when the altitude grows. So there should be a good match between the z-ranger measured height and the LPS height at this transition, the transition is arround 1m if I remember well (we are going to update the zranger deck with a new ST sensor that works to much higher distance, at least 2m).

Where are you seeing the instability, is it closer to the ground or around 1m altitude?

Hi Arnaud and thank you for your reply.

As you mentioned, there are problems when the crazyflie reaches the height of around 1 m. I found this: (ESTIMATOR=kalman
CFLAGS += -DUPDATE_KALMAN_WITH_RANGING -DKALMAN_DECOUPLE_XY) on the wiki, should i recompile the firmware or it's already scripted in there?

Thanks.