Re: [Jsbsim-devel] JSBSim Altitude Hold Autopilot example

[Jon S B. <js...@ha...>]

  David Culp <dav...@co...> wrote:

>Would it be useful to have a seperate component to model a PID
>controller?  It would clean the xml up a bit.  It would combine the
>present INTEGRATOR, PURE_GAIN, and SUMMER components used in the
>autopilot.

That would be a PI controller, then. :-)  I suppose it might be good 
to have a monolithic PI[D] controller, but I like the building blocks 
idea better coming from a perspective of control theory books and 
block diagrams - I think it's a better learning tool that way, and 
easier to grasp the whole concept, but that's just me; I'm certainly 
open to suggestions.  I've got something else in mind, as well that's 
related to this and the second order filter ...

It's interesting that I played with a derivative control action in 
this altitude hold module and got unsatisfactory results (bad 
instability at almost any setting). The PI controller was holding so 
well that there really wasn't anything useful I could see the 
derivative control action contributing.  Maybe I am missing something 
here ... ?

Jon