Re: [Flightgear-devel] Generic Autopilot Rework Proposal

[Josh D. <jos...@gm...>]

Hi James,
Go ahead! I would also study you change as I am interested to learn to
contribute more here :)

Hi Pavel,
This is not a DC-8, actually this demo plane only exists because back in
2020 when I was developing the MD-11's new FADEC/Engine simulation, I
wanted to see how scalable the calculations were. So I attempted to model
the 717's BMW Rolls Royce Engines because the logic used between the MD-11
and 717 is almost identical. So I had a FADEC and engine file sitting
around for it (but no FDM). For this demo I wanted an aircraft I knew had
good flight dynamics, but in those planes everything is so integrated that
it can be tricky to strip out the AP and place the generic back. So I just
paired up my MD-80 aerodynamics, a generic FCS, and 717 FADEC for this
demo. Other than flying characteristics, it's pretty much unusable, so I
don't going to push it anywhere. If you are interested in flying something
similar, you may enjoy flying one or both of my MDs:
https://github.com/Octal450/MD-11
https://github.com/Octal450/MD-80

Also, RE time acceleration, this is a limitation of your processor not the
AP, the autopilot is scaled up in time (because he is operating at FDM
rate) but your processor may stop being able to compute the things well. In
my custom APs, everything is tight tuned to the plane its flying, so if you
have powerful processor, you can time acceleration more. I have a
overclocked Intel 4790k, and can do 8x with functioning AP, sometimes even
16x. But because generic is a generic, it will be a little less precise
than a custom one at time acceleration but I've not tested this yet.

Thanks very kindly for your feedback! I will continue getting the demo
ready, and await the change.

Kind Regards,

--
Josh Davidson


On Tue, Jan 11, 2022 at 7:46 AM Benedikt Hallinger <be...@ha...>
wrote:

> James, that would be nice, I already have some use case in mind :)
>
> Am 2022-01-11 9:44, schrieb James Turner:
> >> On 10 Jan 2022, at 17:50, Josh Davidson
> >> <jos...@gm...> wrote:
> >>
> >> Thanks. I will check these out - my C++ is a rusty tho, but I agree
> >> the first option would be a good feature to have.
> >
> > It’s a very easy change on the C++ side, so I am happy to do it, for
> > the sake of a more flexible AP (and indeed, for encouraging more use
> > of property-rules, since they are far-and-away are most efficient
> > solution for a whole range of stuff)
> >
> > Kind regards,
> > James
> > _______________________________________________
> > Flightgear-devel mailing list
> > Fli...@li...
> > https://lists.sourceforge.net/lists/listinfo/flightgear-devel
>
>
> _______________________________________________
> Flightgear-devel mailing list
> Fli...@li...
> https://lists.sourceforge.net/lists/listinfo/flightgear-devel
>