Bad Positioning Effect in Some Area

All discussions related to the Loco Positioning system
Post Reply
Sun
Beginner
Posts: 9
Joined: Wed Jun 06, 2018 2:50 pm

Bad Positioning Effect in Some Area

Post by Sun » Tue Aug 07, 2018 5:02 am

Hi crazyfliers,

I establish lps system using 6 anchors in TWR mode. I let the drone fly through four points by autononousSequency.py. The sequence I tested is (1.0,1.0,1.0) -> (2.5, 1.0, 1.0) -> (2.5, 2.5, 1.0) -> (1.0, 2.5, 1.0) in unit meters. The locating performance of the drone is excellent at point (1.0, 1.0, 1.0), but bad at other points. Then I find that the drone can stay perfectly at (1.0, 1.0, 1.0) for a long time. But for other points. the drone can't even stay there! It oscillates greatly in the range of about plus/minus half a meter around that set point.

I changed the motors to 720 motors and the battery to 380mah before. I also increased P and D value for this.pidVX, this.pidVY, this.pidVZ and this.pidX, this.pidY, this.pidZ in position_controller_pid.c. The new parameters improve the performance but the drone still oscillates in most areas. The PID parameters may be needed to adjust more. But since the drone can stay at some points perfectly, I think the current pid parameter should ensure the drone to be stable at any point. Some other problems must exist.

The computer with crazyradio is placed near the origin on a desk.
The positions of anchors are listed below in unit meters and they all work normally during the flight.
anchor 0 : 0.04 0.09 0.26
anchor 1 : 0.13 3.80 0.26
anchor 2 : 5.63 1.91 0.26
anchor 3 : 5.67 0.00 2.96
anchor 4 : 0.10 2.03 2.96
anchor 5 : 5.62 3.73 2.96

I will be grateful if you could help me find out the reason of this problem.

Thanks.

arix
Member
Posts: 40
Joined: Fri Sep 01, 2017 1:45 am

Re: Bad Positioning Effect in Some Area

Post by arix » Thu Aug 09, 2018 12:58 am

I use 8 anchors and try tdoa2 twr and tdoa3, i have same problem with you. Your drone can hover at some position. My drone can't hover at all. It always drift. And i try to change position pid,but also not have good hover. However, if i use add flow deck, it have good hover, but a while the kalman is drift.

Sun
Beginner
Posts: 9
Joined: Wed Jun 06, 2018 2:50 pm

Re: Bad Positioning Effect in Some Area

Post by Sun » Wed Aug 15, 2018 11:28 am

After some test, I think I solved the problem.

Mainly I have done two things. One is flashing the newest firmware to the drone, which gives better control of the drone. The other is replacing antennas of all the lps anchors by better ones. This makes the plane hover perfectly at every position in the room.

The next thing should be flying drones with tdoa way.

arix
Member
Posts: 40
Joined: Fri Sep 01, 2017 1:45 am

Re: Bad Positioning Effect in Some Area

Post by arix » Fri Aug 17, 2018 2:24 am

Sun wrote:
Wed Aug 15, 2018 11:28 am
After some test, I think I solved the problem.

Mainly I have done two things. One is flashing the newest firmware to the drone, which gives better control of the drone. The other is replacing antennas of all the lps anchors by better ones. This makes the plane hover perfectly at every position in the room.

The next thing should be flying drones with tdoa way.
Hello i want know you how to replace anteenas ? you don't use dwm1000?

Sun
Beginner
Posts: 9
Joined: Wed Jun 06, 2018 2:50 pm

Re: Bad Positioning Effect in Some Area

Post by Sun » Sat Aug 18, 2018 2:29 pm

Hi arix,
The node I use is a bit different from the one that can be bought from crazyflie shop. I use dwm1000 qfn chip on the PCB board so I can replace the antenna. The change of antenna can enhance the positioning effect but the drone should be able to hover using worse antenna. I think there should be some other problems with your system.

Post Reply