Wind vector from GPS flight data??

[Ralph Weaver]

I'm devloping a GPS data logger for RC planes (www.magtechinc.net). I'm wondering if it's possible to determine the approximate wind vector from a GPS flight log.

If I can determine the wind vector, then I can determine the air speed rather than just the ground speed.

I'm thinking that the average gound vector for the flight would be approx the wind vector.

Does this make sense? Is this possible?

[mnowell129]

Why not just measure the airspeed directly?

[Ralph Weaver]

How would you measure it with a GPS?

[mnowell129]

not with the gps, with an airspeed indicator, there are several commercial ones available.
These guys make one :
http://www.eagletreesystems.com/UAV/uav.html

[jkettu]

Mnowell is right, I don't see any way to get airspeed or wind vector from gps data alone. Computers in full-size gliders take airspeed from pitot-static system, groundspeed from gps and compute head/tailwind component from these. To compute the actual wind vector they would need to know also the heading of the aircraft, but none of the computers I am familiar with does this.

[biber]

Some full size gliding computers do that by detecting the circling and evaluating the drift from one circle to the next over a period of circles.

biber

[Brandano]

Actually, assuming you can log the bank angle as well and assuming only coordinated level turns, you could infer the approximate wind vector based on the drift from the standard turn radius for a specific bank angle and speed

[mraero]

Typically wind speed and direction is computed by comparing three sets of quantities: ground speed and direction (from GPS), heading (from a magnetic compass or similar), and true airspeed (from an air data system, i.e. pitot tube).

There are flight test based techniques for computing wind speed, some of which have been mentioned already in this thread; however, flying a set of precise maneuvers on each flight to determine wind speed may not be practical for your application. In fact, most of the maneuvers are actually designed to eliminate the influence of wind on flight test data rather than to compute its magnitude.

If you really want to log wind speed and direction, you may want to invest in a complete air data system to augment your GPS.

Best of luck!

[HugePanic]

you should keep in mind that the wind vector is different between ground and in flying-heigt...

[Ralph Weaver]

Thanks for the input.

I have a GPS data logger for RC planes. It does not have the option to add airspeed. I know that I can get airspeed if I buy an airspeed logger.

I also know it can be done. The software "SeeYou" is for full scale competition sailplanes and it calculates the average wind vector from only GPS data and it's pretty accurate.

Since I'm only interested in aproximate values and this data would be primarily of interest for RC sailplane use, I think I can get the results I need. My first attempt will be to get the average speed while the plane's heading is between 0-30deg, 31-60deg, ... etc. I suspect I'll have to filter the input data to eliminate anomolies. We'll see...

[Brandano]

Oh, but in a sailplane you can infer a lot of stuff from attitude alone. An airplane has an engine messing up your values.

[kbosak]

Some full size gliding computers do that by detecting the circling and evaluating the drift from one circle to the next over a period of circles.

biber
http://williams.best.vwh.net/avform.htm#Wind
aviation formulary, section "TAS and windspeed from three (GPS) groundspeeds."

[Ralph Weaver]

Thanks, that's the sort of stuff I was looking for. I think that i'll need to filter out any data that includes major altitude changes or very high or very low speeds, then average the ground speed in each of the 3 quadrants? tridants?