How to use Astropy to calculate pixel positions of targets for an arbitrary (imaginary) telescope? (to generate test values for other software)

I’m hoping to use astropy to generate some values to use as truth values for testing some code in a simulation framework I’m working on. Specifically, I’m hoping to test the pixel coordinate calculation I’m writing in the software for both equatorial-oriented and horizon-oriented fields of view.

Looking at the documentation, and specifically the example at First Example — Astropy v5.2 , it looks like I will need to create a WCS object and set various FITS-header-related parameters on it. The problem is, in my research astronomy days, I never did much with FITS files beyond just read them in and then process the images, so I don’t know what the appropriate values are to give to the various attributes (crpix, ctype, etc.) for my situation. I am able to specify an arbitrary size for the field of view, and arbitrary pixel size, in my simulation framework so I have flexibility there.

The primary hiccup is that my simulation framework doesn’t have a direct representation of RA/dec among its available coordinate systems. What I do have available to me is the typical WGS-84-reference lat/lon/altitude, WGS-84-referenced Cartesian Earth-centered Earth-fixed coordinates, and GCRF coordinates (though support for IERS Earth orientation parameters is somewhat limited, so if GCRF gets used on the astropy side I would prefer if Earth orientation parameters can be turned off in astropy, if possible).

So I’m looking for a way to convert from one of the above coordinate systems (and prefer one of the WGS-84-based ones over GCRF because of the nature of the framework I’m working in) to pixel coordinates given an arbitrary sensor location, pointing direction, and field of view for both equatorial-oriented and horizon-oriented fields of view. I’d be equally happy with the reverse conversion, pixel coordinates to one of the above coordinate systems, since that can provide me with usable test values equally easy.

(When I say “arbitrary sensor location, pointing direction, and field of view,” I do intend to fix those values for generating the test numbers, probably over a small range of values.)

Also worth noting that the telescope will be looking at things in Earth orbit, so an assumption of things being infinitely far away won’t hold.