So I'm assuming you've built this new dimension in either a development or test environment, certified that it does what you need, and now you want to deploy this to your production environment. Right?

Now I'm hoping that when you say your DSV is different you're just talking about the ID. If so, you can change the XMLA to use a parameter $(DataSourceViewID) in place of your DSV, then when you run ascmd you'd execute ascmd.exe -S server_name -i my_file.xmla -d db_name -v DataSourceViewID=my_id

But when you say everything would be different in prod, you could be talking about a much bigger problem. If your analysis database is very different between dev/test and prod (not just IDs but column binding definitions, etc.), that's something you should work to correct. For this project perhaps go the route of a document describing all the setup steps, but you'll want to script your production cube and bring it back to your test/dev environment(s) so you don't have this problem again.