The files on any crystallographic database are real samples, subjected to any effect due to isomorphic or interstitial substitution, particle size, strain, or another sort of effect.
Therefore, you may check the results of your work with any free available database, such as the American Mineralogist Database (http://rruff.geo.arizona.edu/AMS/amcsd.php), Crystallographic Open Database (http://www.crystallography.net/cod/) that will correspond to a natural or synthetic sample deposited by the authors.
There is no available database with only synthetic pure standard materials or mineralogical phases.
The phase you are dealing with may present several effects based on Al substitution (amount and site) in the spinel structure.
Any magnetite CIF file can be used as model to start the analysis of your result. The main difference will be the lattice parameters and the amount of iron substitution, the occupancy. Iron is much heavy and will dominate the diffraction.
Consider the effects already visible in the CIF files used as references, may not correspond to the same you may identify in your sample.
I expect that having substituted a part of iron with aluminum the structure of magnetite could be maintained. So you can use the PDF card of Fe3O4 to verify the peak positions, then to determine lattice parameters, crystallite sizes, occupancies and so on you can perform a structural refinement.
In addition to the structure of Fe3O4 magnetite, as suggested by @Marcella Bini, you can start from the structure of FeAl2O3 hercynite or from some member of the isomorphic series. The data can be downloaded, for example, here: http://rruff.geo.arizona.edu/AMS/amcsd.php
The effect of composition on the diffraction pattern is very low for spinel phases. Evaluation using ICDD data is an unrealistic matter for them.
In the end, Rietveld's refinement also does only little sense. The reason is too great the possibility of isomorphic substitutions. All atoms are in special positions. Vegard's rule for spinel phases does not work or only in a limited extent. Especially with natural samples, it's bad.