Re: [Jsbsim-devel] New Trim and State Space Routines Added

[Ron J. <wi...@je...>]

I think for now I will push a fix to the property names in FGFDMExec. As it 
sits right now the defaults can't be set from a script or the command line 
since the properties they depend on don't exist. Also, I'm going to make sure 
everything that's read gets set.

Ron

On Sunday 28 October 2012 15:51:32 Jon S. Berndt wrote:
> I noticed that, too - but not the different property names between the two
> files. I merged the trim code into the current JSBSim with very few
> alterations. Agostino and James will be looking at the code in the next few
> weeks.
>
> I have requested that consideration be given to the use of a configuration
> file instead of properties for trimming.
>
> Jon
>
> > -----Original Message-----
> > From: Ron Jensen [mailto:]
> > Sent: Sunday, October 28, 2012 3:14 PM
> > To: Development issues
> > Subject: Re: [Jsbsim-devel] New Trim and State Space Routines Added
> >
> > On Wednesday 24 October 2012 23:05:39 Jon S. Berndt wrote:
> > > I finally got around to adding James Goppert's trim and state space
> > > classes and routines to JSBSim CVS.
> > >
> > > Jon
> >
> > Ah hah! I see a problem. In src/FGFDMExec.cpp you initialize all the
> > simplex defaults into trim/guess/*, but
> > src/initialization/FGSimplexTrim.cpp is trying to load the defaults
> > from trim/solver/*.
> >
> > Going on from there, I'm seeing the defaults are bad. The fcs needs to
> > be in a
> > 0-1 range instead of a 0-100 range, the alpha and beta angles are being
> > interpreted as radians (at least in some places) and are entered in
> > degrees.
> > Changing that gives me a solver that doesn't move the controls, though.
> >
> > :(
> >
> > I never looked at James Goppert's original code, though. Not sure what
> > else needs to be adjusted.
> >
> > Ron
> >