Re: [Paparazzi-devel] Trajectory pivoting in flight plan and wind direction/speed

[Martin Mueller]

Hi Roman,

> First one: My flight plan trajectories are defined and each of them include a 
> lot of points. I am looking for the way to pivot the trajectories including all 
> the points about one point, so that the first part of the trajectory is always 
> directed against the wind flow direction. Of course the wind direction will be 
> defined before each flight. So I need to change all trajectories as I described 
> before. Have anybody used something like this? Have you any ideas how to do it?

have a look at nav_cube.c. It contains some trigonometric functions to 
turn predefined lines/points around a center point.

> Second one: I need to define on some way the direction and speed of the wind in the 
> altitudes my flights carried out. As I can see GCS show me some data about wind speed and 
> direction. But I ask me how precise is it? And what methods will be used to calculate it? 
> I had an idea to calculate the wind speed and wind direction using ground speed (GPS), 
> truth airspeed (Airspeed sensor) and drift distance, so as it was done in the "stone 
> times" by the pilots. Another idea is to use a balloon with an anemometer, but the 
> models I could use together with a balloon are quite expensive. Any ideas for how to 
> measure the wind speed and direction in different altitudes using paparazzi equipment? Or 
> some alternative methods?

The wind estimation is done in the gcs laptop with wind.ml. See that 
OCAML file for some theory and links. It uses the ground speed and 
statistic methods assuming that the true airspeed of the aircraft 
remains fairly constant throughout the measurement. The comparison with 
a tethered ballon showed that it is pretty accurate

http://iopscience.iop.org/1755-1315/1/1/012014/pdf/ees8_1_012014.pdf

There is another approach in the MatrixPilot that might be interesting 
to try

http://gentlenav.googlecode.com/files/WindEstimation.pdf

http://www.google.com/codesearch/p?hl=en#_3qlLx8yAec/trunk/libDCM/estWind.c&q=wind%20package:http://gentlenav\.googlecode\.com&sa=N&cd=1&ct=rc

Would be great if you could implement the classic approach with the 
airspeed, groundspeed and maybe a magnetic sensor.

Martin