The thesis introduces a new method for determination of equivalent electromagnetic parameters of conducting materials considering the losses due to alternating magnetic field. The method combines evolutionary optimization algorithm and calculation by finite element method. The differential evolution determines the equivalent parameters of the material which in finite element calculation yield the same losses in the volume of that material as the measured ones. The toroidal solid cores and rods were made for measurement of static magnetization curve and specific electrical conductivity of nonmagnetic and magnetic steel. Thus obtained material parameters were used for determination of their boundary values for calculation of the equivalent material parameters by differential evolution. The coil for direct measurement of additional losses was made. A series of measurements on various samples of different materials were conducted. The 3D model of the coil was made in finite element software and using the differential evolution method, the equivalent material parameters were determined. The calculation was compared with measurement. Using a 2D rotational symmetric model of the plate with the thickness of 10 mm it was analyzed how the material parameters affect the amount of losses in the sample placed in the sinusoidal magnetic field of constant amplitude. It was shown that losses have the bell-shaped dependence on material parameters due to the fact that losses tend to zero when the conductivity of material tends to zero or infinity. Using differential evolution the material parameters for which losses have a maximum value were determined. Three models of transformers were made. The equivalent material parameters determined for the coil were applied on those models in order to determine using finite element method the share of hysteresis losses in the total amount of losses in structural parts of a transformer. The largest possible losses in structural parts of transformers that can be calculated using finite element method were also determined. The results were compared with the measurement and the reasons for discrepancies were explained.