Reply to: Re: [Paparazzi-devel] distance measurement for landing.(Bernard Davison)

[Chris]

>But they can be easily integrated with I2C units which are available.
>These could give a general reference which then hands over to
>another system that is more accurate such as the ultrasonic or laser 
units.

That is exactly what i have done.
I use a baro (scp1000) that gets calibrated every minute from
many averaged and matched gps readings when the plane flies level only 
of course.
The barometer is the main altitude instrument but as i said it gets 
calibrated often from the gps.
This way the barometer not only provides "accurate" altitude but also 
very accurate ascent and descent rates.
I have also increased the vertical and horizontal control loops 
frequency from 4Hz to 10Hz since the barometer's refresh rate
permits this and in simulation the results are fantastic so far.
If the ultrasonic range finder detects with certainty that we are below 
5m or whatever altitude gives good readings and if the ultrasonic range 
finder is also enabled from the flight plan, it takes over for the landing.
I have implemented strong filtering with prediction methods so
the ultrasonic range finder does not get tricked from noise.
Btw the altitude from the sonar id fed to the kalman filter for now.
Of course only testing will show if i have done more than a hole in the 
water...
The ultrasonic range finder is here so i will test it next week.
Chris