Pseudopotential is a concept related to ab initio calculations, not molecular dynamics. It is based on the assumption that in chemical reactions, only the valence electrons are involved. Therefore, treating the core electrons as an effective potential instead of individually drastically decreases the computational time of the calculations.

You are right in that EAM/MEAM are suitable to describe the FeCo alloy. Once you have the EAM/MEAM parameters for each metal element, those of the combination can be done automatically by LAMMPS with the command

pair_style eam/alloy

Here is a hint of how to do so:

http://lammps.sandia.gov/doc/pair_eam.html#pair-style-eam-alloy-command

The same can be applied for a third and four metal element. 

However, if the other elements are non-metallic, things get nasty. You may use Lennard Jones potentials in combination with EAM/MEAM for which you can use this link:

http://lammps.sandia.gov/doc/pair_hybrid.html

Still, this implies some approximations since LJ is a pair potential, therefore, the energy of two atoms is only a function of their distance, neither their bond order, nor the particular structure that contains these atoms.

Hope it helps!

Thank you G. J. Soldano for your quick answer.

In fact, I habitually used "pseudopotential" name instead "potential" (in my everyday work I'm dealing more with DFT than MD).

Can you suggest me where can I find the EAM/MEAM parameters for my alloy elements?

I agree with everything that G.J. Soldano discussed.  I would also add that there is a piece of software called GULP that will do some automated fitting of potential parameters to input parameters (maybe experimental or ab initio values of lattice constants, energies, elastic constants, etc.).  I have used this code to make force fields of my own if you can't find something suitable already published:

http://www.staff.amu.edu.pl/~magchem/GULP/gulp.html

Dear K. Gruszka,

In this link you can find many corrected EAM parameters for many atoms and several combinations.

https://sites.google.com/site/eampotentials/

I have talked to some colleagues who are closer to this field and they tell me that the FeCo is not well described with EAM/MEAM.  Apparently, many physico-chemical properties of this alloy are strongly dependent on the structure, geometry, etc, which is hard to contemplate with standard potentials...

I attached you a first principle work on FeCo, which you probably have already, but just in case.

Good Luck!

Thank You all for answers.

J. Trinastic: I think that this GULP software is rather for crystalline materials, not amorphous one - like in my case.

J. Soldano: I've seen this site before, but I can't find it helpful because my system is quaternary.  My mistake that I didn't precisely write what kind of alloy I'm trying to simulate.

The matter is that I want to simulate annealing (melting) alloy and then do rapid solidification. Composition is based mainly on Fe element, but not only; the exact composition is following: Fe62Co10Y8B20. This is my basis which I want to dope with 1% Cu at. in expense of Fe 

Up till now I've seen many binary and ternary systems, but as You can see my is more complicated one.

So, are EAM/MEAM potentials good for my case?

You can and also Gupta potential

I agree with S. Nouemo Gupta potential can be use in this case