In power system, the application of the power system stabilizer (PSS) has been proved to lead to the stability of power system. It is necessary to determine the appropriate PSS parameters. In this paper, a new cost function is presented for the tuning of PSS parameters. Optimization of PSS parameters has been done by Particle Swarm Optimization and Imperialist Competitive Algorithm. The power plant system that is simulated in this paper is one machine to infinite bus system. The responses are compared with CPSS. The responses PSO-PSS and ICA-PSO are also compared with each other and the efficiency of the new objective function (SMVS-DFT) will be proved.