Really, don't do it.
I guess it applies in general, but specially if you write scientific research papers, the Gratuitous And Excessive Use of n-Letter Acronyms is really bad for you.
The use of acronyms doesn't really help you much in saving space to say more interesting things in your paper (and there are better tricks you can do for that). But the use of acronyms will definitely make your paper more difficult to read.
Specially if you work in a very narrow area of research (and most of us do!), anybody outside of your little area of interest will have no idea of what you are talking about.
So, unless the title of your paper is “P ≠ NP” (or maybe “P = NP”), avoid the use of acronyms in the title and abstract of your paper. You will only manage to lure away readers which might otherwise be interested in your paper. And you want more people to read your papers, not less!
Also avoid acronyms in the body of your paper as much as you can. And, if you have some really good reason for wanting to use an acronym, make sure that you define its meaning the first time you introduce it.
(Yes, I’ve made this very same mistake a few times in the past, but the amazing thing about human beings is that we are able to read and consider a good argument, learn from past experience, and correct our own behaviours)