lighthouse signal is unstable

Topics related to the Lighthouse positioning system, configuration and use
Post Reply
yuwen
Beginner
Posts: 2
Joined: Thu Oct 28, 2021 11:37 am

lighthouse signal is unstable

Post by yuwen »

Hi,

I am new to crazyflie. I use two lighthouse V2 base stations for positioning and try to use LH to autonomy fly. After setup all equipment, I use LH tab in cfclient to estimate the geometry and see if the position is reasonable. When the crazyflie is near to the origin point, the position is correct. But when I move it a little far, the position starts to drift quickly and the receive indicator turns red(I think means the crazyflie can't receive the signal from the base station).

The range for the stable signal is only 2.5m*2.5m and 1.7m in height, but I expect the range should be much larger(about 4.5m*4.5m and 3.5m in height). I set two base stations in 3.5m height and place them in diagonal with a distance of 4m. I have tried changing the tilt angle of the base station, horizontal, 45 degrees, face down, but doesn't get better.

Is the range the limit of LH? or any adjustment in hardware, software or environment can make it better?
kristoffer
Bitcraze
Posts: 630
Joined: Tue Jun 30, 2015 7:47 am

Re: lighthouse signal is unstable

Post by kristoffer »

The maximum range of a V2 base station is around 5-6 meters, but the actual space where the position can be tracked also depends on the geometrical setup of the system. The light sweeps from the base stations are detected by the four sensors on the top of the lighthouse deck and if the angle to a base station is too shallow it will not be detected, for instance if flying close to the same height as the base station you will probably not get any data. The same will happen when flying way from a base station or when turning and the CF is tilted away. You can use the client to see which base station that is received and figure out how your geometry works out.
Note; If you are using the standard set up, reception of one base station should be enough to fly.

The base stations have a quite wide "field of view", 120 degrees in one direction and 160 degrees in the other I think so usually you only need to point them in the right general direction. I believe the calibration error is smaller in the center though and it is probably a good idea to keep the main flight area in this field.

It is hard to say anything specific about your set up but the higher the base stations are place the bigger volume will be covered.
When the crazyflie is near to the origin point, the position is correct. But when I move it a little far, the position starts to drift quickly
Do you mean that the estimated position has a large error when you move away some distance? If everything works properly I would expect the error to be less than 5-10 cm. If this is the case, are you sure that you had received the calibration data before estimating the geometry?
yuwen
Beginner
Posts: 2
Joined: Thu Oct 28, 2021 11:37 am

Re: lighthouse signal is unstable

Post by yuwen »

thanks for your response

I mean, sometimes, when I put the crazyflie 2 meters away from the origin, the lighthouse doesn't work well, and the estimated position is several meters away from the correct position.

I think I will try to adjust the position of the base station to make it better. But now I want to ask some questions:

1. Would the glass window nearby the fly area reflect the IR signal and cause some problem?

2. Can I use more than 2 LH base stations to extend the fly area or make it more reliable.

3. If I use the LH deck and flow deck at the same time, how would they work together?
kristoffer
Bitcraze
Posts: 630
Joined: Tue Jun 30, 2015 7:47 am

Re: lighthouse signal is unstable

Post by kristoffer »

the estimated position is several meters away from the correct position.
That does not sound right at all.

1. Yes reflections is definitely a problem and we have heard of users having problems with windows. Try to cover it to see if it makes any difference.

2. No but a bit of yes. There is experimental support for 2+ base stations, but functionality is missing for estimating base station geometry. Unless you really understand the system it is probably hard to get it to work. I think there are some threads in the forum with a bit of information if you want to try. We hope to add official support sometime in the future.

3. Yes you can use LH and Flow at the same time. They both add information to the klaman state estimator that combines it to one solution. If you fly outside the LH coverage area only the flow deck will be used and everything should work just fine. When you get back in the LH area the estimated position will again be based on both LH and flow. You can check out the hyper demo https://www.bitcraze.io/2020/12/the-hyper-demo/ where we mixed LH, flow and Loco positioning.
Post Reply