I never use MAUD. But for refinement of doped materials just keep the same the atomic fractional coordinate as that of atom in which side you made doping but don't forgot to change the occupancy accordingly. For intercalation just use intermediate coordinates and check for which it will give the best fit.

Since intensity is not only affected by substitution (which varies the effective atomic scattering factor of position x,y,z by different atoms, but also vacancies...) I would not believe a refinement. The problem is simply that you cannot exclude all other effects to cause these intensity variations. Of course you can simply assume that nothing is changing and everything is ideal, but even without vacancies this is an assumption which is not proven. You can simulate patterns and study how much they are changing with the variation of vacancies. In t hat case you simply need to change the fractional amount of atom A and B and C etc  ... for the same position. Sum about A+B+C

Sorry, can you be more specific? What "voids" in the structure? Also, to make a refinement in MAUD you are to provide some structure initially, and then you can tweak it to match the measured pattern; in MAUD it is even possible to plot this structure in 3D right away. So I am a bit confused with this "structure-to generate CIF-to generate struture" sequence.

ya in this case of MAUD  I have to provide some structure parameter initially. During the refinement process I have to change those parameters manually for fitting.after fitting is done we can generate a cif file for that particular material.

And the voids i mentioned there is tetrahedral and octahedral voids. My qstn is due to doping is there any change in those types of voids..