Purpose. The scientific work provides for the development and justification of a new method for the parametric identification of electric traction devices based on a system of orthogonal functions, namely – on the basis of time power series. Methodology. To solve this problem, we use: the theory of power series; basic laws of theoretical electrical engineering, elements of the spectral analysis of periodic and non-periodic non-sinusoidal functions of electrical quantities of voltage and current; matrix methods for solving a system of algebraic equations. Findings. A new method of time power series is developed, which can be used to identify devices both traction power systems and electric rolling stock. The method is based on the compilation of integro-differential equations of the electromagnetic state of the electrotechnical system under study. The given voltage and current in the indicated equations are approximated by power series, into which the required parameters of the electric circuit of the system enter in the form of coefficients. Using the developed method, electric locomotives AC of DSС3 and 2ES5K were identified in the work. Their parameters R and L, varying in time over the period of the applied sinusoidal voltage, are determined. Time dependences R(t) and L(t) are constructed and analyzed. An explanation is given for the regions with negative inductance. It is shown that these electric locomotives can be identified by a passive two-terminal network with parametric resistive and inductive elements. Originality. Innovation of work consists, first, in development of a new method of identification, and secondly, in the possibility of replacing any device of electric traction with parameters of RL- a two-terminal network. And, finally, the fact that the proposed method makes it possible, by improving the approximation of input and output voltages and currents, to obtain the most accurate values of the parameters R and L the identification two-terminal network. Practical value. The developed method makes it possible to evaluate the parameters of devices of electric traction systems on a real-time scale, which facilitates its application in modeling the transient electromagnetic processes that arise during the operation of the system under study in emergency modes.