UWB test for Swarm

Firmware/software/electronics/mechanics
Post Reply
SH_Lee
Member
Posts: 51
Joined: Tue Feb 18, 2020 8:48 am

UWB test for Swarm

Post by SH_Lee »

I'm doing a uwb sensor experiment for swarm.
Location control using loco positioning deck (lps) was successful when there was only one drone.
We received stable sensor values, and control to the desired location was also successful.
We completed an algorithm for one drone and tested it with two drones before flying several drones.
Unfortunately, we didn't get the correct sensor value with two drones.
Both drones have completely turned off the uwb sensor's led
I checked the client's plotter and found that the value remained unchanged and remained very large.
We are using the TWR method and will control a total of 4 drones.
I know how to officially recommend it is TDOA2, but I want to maintain it if possible because I got very satisfactory results when I controlled one unit with TWR method.
If anyone knows about this, please give me good information.
kimberly
Bitcraze
Posts: 1050
Joined: Fri Jul 06, 2018 11:13 am

Re: UWB test for Swarm

Post by kimberly »

Hi!

It is very unfortunate but TWR will not work good for more than one drone. The reason is that with TWR, the drone is the one that start sending the packets to the nodes, and if two of them does that the packets will collide. That is why you are getting a worse result with 2 instead of 3. The drones will need some kind of synchronizing method to make sure that they are not sending packets at the same time, which would be something you would need to implement yourself.

TDOA2 and TDOA3 should work. You can try to tune the kalman filter and the control parameters to get a better performance. Adding a flowdeck should also improves the quality of flight.
Post Reply