Discussions

Drift correction with Magnetometer

Posted by Rupert Kent on November 24, 2010 at 1:55am

Hi chaps,

 

I am planning to add the magnetometer to my APM to provide heading information.

 

As a PPL holder, I am fully aware that GPS can only provide track information and as such, GPS based navigation is inefficient in anything but dead calm conditions (the aircraft will fly in a curve between waypoints).

 

With a compass on board giving heading, it's possible to measure the delta between heading and track to understand drift and subsequently compensate accordingly.

 

My query is what is the additional info provided by the magnetometer used for by default...? Is correction of drift already included in the APM software build...?

 

Thanks in advance,

 

Rup

You need to be a member of diydrones to add comments!

Join diydrones

Email me when people reply –
Follow

Replies

  • Rupert Kent November 24, 2010 at 11:38pm
    Thanks Bill. All interesting stuff. I have a couple of magnetometers on their way and await to see how they can be integrated.
  • Rupert Kent November 24, 2010 at 7:41am
    Hi William. Thanks for the info. That's really neat. I am guessing there's a whole bunch of vector matematics behind the scenes, but this is really nice. I am guessing that the gyros on the IMU Shield are plenty accurate enough to compensate for wind, provided the wind isn't shifting so rapidly (in which case, drift is probably the least of your worries anyway).

    I wonder if this will eve make it into APM...? I am not a programmer and as such have no clue how to implement this, but picturing a non-drift-corrected flight plan in a moderate wind, the gain in efficiency by flying direct must be considerable. 10%, maybe more...
  • Rupert Kent November 24, 2010 at 4:51am
    Thinking about it a little more, this makes complete sense. There's no need for the magnetometer in this scenario. The only time the magnetometer comes good is when the craft is stationary and the GPS cannot provide track information.

    Pretty sure the APM doesn't correct for wind at this time...
  • Ken November 24, 2010 at 4:47am
    You have it a bit backwards.

    Now, I don't use Ardupilot software, I have my own projects and only hang out here for the pleasant company. So I can't comment on whether or not the ArduPilots have any track following.

    But it is entirely possible with a GPS to know your cross-track error (and correct for it provided crosswind/cross current does not exceed vehicle speed.) True, the little modules we use don't (usually) know about such things as waypoints, tracks, etc. so that must be handled by the host processor. The aviation formulary has all you need to find a line from waypoint to waypoint, determine current cross track error from that line, velocity made good, etc.

    By having an assumed (wx forecast) winds aloft, in flight calcs, and optionally a magnetometer, you can round one waypoint and set a new course to steer before the GPS location shows any CTE making things a little smoother.
This reply was deleted.

Activity