Discussions

Loiter Tuning

Posted by Bob Wilkinson on July 9, 2012 at 8:41pm

I flew my 3DR-B tonight in startlingly windless conditions.  It was the first time I've really tried to test Loiter.  I got the quad into a nice stable hover (it was hovering practically dead still in these conditions with no control input at all) and flipped to LOITER.  It did a pretty good job of staying in place but started to sort of orbit around the spot it started in.  These orbits slowly got wider, to the point where I got uncomfortable and switched back to stabilize, though it took a minute or so.

Is this normal?  Do I need to tune my LOITER PID?  I noticed that the simulator has a default LOITER D of 0 while the default for Arducopter is 0.4.  Changing the sim to 0.4 seemed to make it oscillate more in windless conditions, so maybe the answer is to change my D to 0?  But will that negatively effect performance in windy conditions?

Here's my log from tonight: 2012-07-09%2023-07%202.log

Thanks for any advice.

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

Join diydrones

Tags: pid, loiter, tuning
Email me when people reply –
Follow

Replies

  • novrizal herdananto January 10, 2013 at 3:38pm

    hi,

    i've just install 2.9 rc2 on my apm2.5. i have just tried loiter in stock pids, and i see quite oscilations.

    since i'm a noob, i need advice in which PID i should tune ? attached is my loiter log and param.

    2013-01-11 06-25 1.log

    my_param.param

  • Bob Wilkinson August 23, 2012 at 9:02pm

    I had some success with reducing the loiter P, I got it to the point where my loiter would stay within a 5m box or so.

    Then I got the ublox GPS, which is worlds better in terms of locking on quickly and getting a higher satellite count / lower hdop.  However when I did some loiter tests I got a very strange result where the quad started an expanding spiral that rapidly got out of control.  Anyone know why this would happen?  I haven't had much time to play with it.

    Logs from the incident I'm referring to are attached.

    Thanks

    2012-08-23 23-32 3.log

    2012-08-23 23-32 3.kmz

  • Murray Spoelstra July 9, 2012 at 10:14pm

    I found that with my small quads it works best to reduce Loiter P to 0.2 and Loiter I to 0.05. It also seems to make a difference when I change Nav P to 0.9 and Nav I to 0.05. Finally got Loiter and waypoint flying OK in wind with purple motors, 10x4.7 APC props, 1250g weight and motors on 570mm pitch diameter.

    Problem I have though is that when I fly waypoints and choose Loiter turns, it goes into a 45deg bank and drops out of the sky when it starts executing the command. Any suggestions please?

  • Developer
    Jason Short July 9, 2012 at 9:41pm

    What your seeing is lag in your GPS input being too much for the rate_P value you have. Try lowering loiter rate P.

    The reason it's jumping out of a perfect loiter is the GPS hits a threshold and makes the quad think it's traveling faster than it is. it over-corrects with the High gain and you start an oscillation.

    Jason

  • 3D Robotics
    Chris Anderson July 9, 2012 at 8:54pm

    What version of the code?

This reply was deleted.

Activity