In order to maintain equilibrium in small or large tokamaks poloidal field coils are utilized, since the function of the poloidal magnetic field is a complex function of current density and the position of the coils, a change in any of the parameters can have a strong effect in the confinement and the magnetohydrodynamic parameters. On the other hand, considering the continuity of the current and the position of the coils, the space being searched is so big that taking all possible conditions into account becomes practically impossible. So a method should be utilized that is able to optimize the position and current of the coils without searching the whole space. This paper seeks to find a new method of deriving the plasma parameter in which a combination of the two methods of neural network and Particles Swarm Optimization is used in order to optimize the position and current of poloidal field coils in Damavand tokamak. Since in the employed methods no special topology is applied, it can be readily used to study any other tokamak.