If you're interested in the structure at 30 nanometers, are you not better served by looking at the data in reciprocal space?
To fit PDF data out to 300 AA, you need (1) a quite high resolution diffractometer, and (2) a good understanding / model of resolution effects in the data (for instance, as discussed here https://doi.org/10.1107/S2053273318003224
If you mean that you wish to extract a size estimate for ca. 30 nm particles, your problem is to distinguish between size damping in direct space and resolution damping, both of which are multiplicative gaussian(ish) envelopes (with a bunch of assumptions rolled in).
Perhaps I'm not reading the question correctly, but it seems that at 30 nm you've left the domain in which PDF excels, and would be better suited by a return to reciprocal space.