Identification of the constants of material models is always a concern. In the present work, a combined experimental, numerical and optimization technique is employed to determine the constants of Zerilli-Armstrong model. The experiments are conducted on a compressive Hopkinson bar, the simulations are performed using finite element method and optimization is carried out using genetic algorithm. In the method adopted here, there is no need for experimental stress-strain curve which is always accompanied by restricting phenomenon such as necking in tension and bulging in compression. Instead of stress-strain curve, the difference between the post-deformation profiles of specimens obtained from experiment and the numerical simulations is adopted as the objective function for optimization purposes. The results suggest that the approach introduced in this work can substitute costly instrumentations normally needed for obtaining stress-strain curves at high strain rates and elevated temperature.