Is it possible to control crazyflies via collective thrust and bodyrates (CTBR) ?

[Sapio-S]

Hi!
I just read a paper that says controlling drones via CTBR can be more robust than setting linear velocities or single-rotor thrust. However, I only found CmdVel (or cmd_vel_legacy) function in this repo, which takes [roll, pitch, yaw rate, and thrust] as input instead of [roll rate, pitch rate, yaw rate, and thrust]. I also searched the cflib documentation and the crazyflie firmware documentation but couldn't find any function that takes in the latter form of input. (CPPM Emulation looks promising, though. ) So here are my questions:

• Is it possible to control crazyflies via CTBR? Do I have to change the codes of crazyflie firmware, cflib, and Crazyswarm1/2 to send CTBR commands?
• Can I somehow convert the CmdVel / cmd_vel_legacy command into a CTBR control command?
• Does crazyflie have a controller that takes in body rates & thrust and outputs rotor commands? If so, where can I find the controller's code? Can the attitude rate controller do the job?
• Alternatively, can I control a crazyflie by sending single-rotor thrust? I am considering bypassing all controllers and taking direct control of all 4 rotors.

Thanks in advance!

[whoenig]

All these questions are best to be asked in the official discussion from bitcraze.

My (incomplete) knowledge: It should be possible with the "legacy"/rpyt mode, see https://github.com/bitcraze/crazyflie-firmware/blob/master/src/modules/src/crtp_commander_rpyt.c. Basically, there seems to be a parameter to switch it to "rate" mode, then the cmd_vel_legacy commands are interpreted as rates, rather than angles. The PID controller should be able to handle this case, the other controllers do not.

Commanding rotors directly is currently only possible via parameters (good for testing, not for real-time control). In any case, you wouldn't be able to fly in a stable manner, due to the communication latency.

[Sapio-S]

Thanks a lot! I will give it a try.