Bitcraze Forums

Hello!

I test 10 crazyflies swarming.
when 10 crazyflie are on swarming, some crazyflie crash or lose balance at the same time.
the timing of these event is random.

movies: (Google Drive)
movie1
movie2
movie3

this is the flight area environment.
the red area is the flight range of the crashed or unbalanced crazyflies
my questions
1. why did the crash and unbalance happen at the same time?
2. How can I know the cause of the problem?

Regards,

Hi!

Thanks for the videos!

It is a bit difficult to see where these crashes come from... It can be perhaps due to bad batteries, wind of the propellers of the other drones or perhaps you are flying to close to the top of the lighthouse basestations. But let's figure it out!

Here are some questions.

- Do you know which controller are you using? PID or Mellinger?
- The bright open door in the back, is that direct sunlight?
- Are the crazyflies running on the latest release of the firmware?
- What are you using to control them? A pythonscript, crazyswarm or are you using the app layer.

Hi, kimberly!
Thanks for reply.

The output of crazyradio-pa is set to the lowest setting (P_M18DBM)

1. I use high_level_commander and upload polynomial trajectory.
2. The video was taken in 15:00~19:00. it is sunlight and fluorescent light. 3. all crazyflies and lighthouse decks are flushed the 2021.03 firmware from cfclient
4. I used crazyswarm of crazyflie-lib-python to control them

Thanks for your answers!



1- I actually mean another controller, the high level commander is the setpoint handler, the controller is what make sure that those setpoints are met. Here is the documentation to explain the whole stabilization loop..

You can check this with the cfclient by going parameter and checking stabilizer.controller, but this could also possibly be switched in run time by crazyswarm... Here are some instructions to try the PID controller in Crazyswarm

2- That should be okay but it might improve if you close the door.

3- Good!

4- So I'm not exactly sure if the crazyflie-lib-python package is used by crazyswarm directly. But did you mean that you use Crazyswarm OR crazyflie-lib-python? In which situations are you observing your problem?

I'm sorry for my misunderstanding.
I don't use CrazySwarm and only using crazyflie-lib-python.

Alright! That is good to know.

So probably you based your script on the hl-commander-swarm.pyor synchronizedSequence.py? If you still have anywhere in the python script where you use activate_mellinger_controller, either comment it out or put it to False.

Then try to fly again and see if it improve anything.