4
$\begingroup$

I want to implement my own GPS navigation for a quad-copter. I can calculate and filter the GPS coordinates (latitude and longitude in degrees).

I believe the easiest approach for me would be, to calculate the change of the heading of the quad-copter from the current attitude to the destination point and let it fly straight on after turning.

However I am not sure about the 2D representation of the latitude/longitude-GPS coordinates (for a round earth to a 2D map system when calculating the heading change). How big is the expected error? Or is there none?

Improve this question
$\endgroup$
1
  • $\begingroup$ Have a look at world geodetic systems (especially WGS84). Also, on a somewhat related note, you can use USGS's earth explorer portal to download anything from satellite imagery to digital elevation models (free registration required though). $\endgroup$
    – EDDY74
    Commented Mar 27, 2014 at 6:03

1 Answer 1

Reset to default
1
$\begingroup$

If the expected range of your quadcopter's journeys is under 10 miles, the error between spherical and flat geometry is less than the GPS's poistion error...assuming you are using normal GPS, not dGPS. As a rough rule of thumb, 2D mapping works to the horizon!

Improve this answer
$\endgroup$
4
  • $\begingroup$ Very interesting way of thinking about GPS errors and errors induced when converting between spherical/flat. Could you however perhaps elaborate a bit more on what you mean by 2D mapping works to the horizon. $\endgroup$
    – EDDY74
    Commented Mar 27, 2014 at 6:09
  • $\begingroup$ I was just asking this question myself yesterday and was thinking how to calculate the error. $\endgroup$
    – dgrat
    Commented Mar 27, 2014 at 18:58
  • $\begingroup$ Sorry, the horizon remark was in the nature of a joke: the horizon is 'visible' because its where the curvature of the earth has become 'visible' $\endgroup$
    – Mhz4.77
    Commented Mar 27, 2014 at 21:32
  • $\begingroup$ I believe this page covers the problems and answers in much greater detail than I ever could: cs.nyu.edu/visual/home/proj/tiger/gisfaq.html In particular, note Bob Chamberlain's message on range errors part way down the page. $\endgroup$
    – Mhz4.77
    Commented Mar 29, 2014 at 13:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.