Re: [Jsbsim-devel] F-16 Stability problems

[Erik H. <er...@eh...>]

I did some work on the PID's and got it almost working but the problem 
is the final details: getting the error to zero.

But the process of starting FlightGear every time to test the changes 
makes it unpleasantly time consuming.

Erik

On 10/20/2016 08:55 PM, David Culp wrote:
> From your description of the problem it sounds like integrator wind-up
> could be the issue.  The fix was added to the PID component a few years
> ago, but I haven't used it yet.  IIRC,  it involves specifying an error
> property that when very near zero causes the PID controller to reset the
> integrator.
>
> Dave
>
>
> On Oct 20, 2016 11:25 AM, "Bruce Clay" <bcl...@gm...
> <mailto:bcl...@gm...>> wrote:
>
>     Erik:
>
>     I realized that my reply had not popped up on the forum yet.  It
>     seems I replied from my work email and that email does not get
>     posted for some odd reason.
>
>     ---- previous reply ----
>
>       In thinking through your response I guess I have I
>     misunderstanding of  the way JSBsim works.  I would have expected
>     the PIDs to be in the base code not in the aircraft dependent
>     scripts.  If we take a simplistic view of a PID as a filtering
>     mechanism how does altitude affect the roll?  The problem tends to
>     show up more in the low altitudes say under 10,000 feet and around
>     600 knots.  Neither of these values are near the limits for the
>     aircraft.  I do not see the oscillations during extreme maneuvers
>     rather just the opposite.  Doing extreme maneuvers seems to
>     stabilize the aircraft (i.e stop the oscillations).
>
>     I am no expert in the math behind the flight dynamics so I may be
>     overlooking something.
>
>     Thanks for your help
>
>     Bruce


-- 
http://www.adalin.com - High performance virtual reality audio software.