Modeling highly nonlinear load dynamics for harmonic assessment

In this paper, an advanced method based on measured data and neural network for highly nonlinear loads such as electric arc furnaces (EAFs) is introduced. An industrial power system with the EAF load is used for field measurements. Modeling results obtained by the proposed method are then compared with the measured data. It shows that the presented method can model the highly nonlinear load dynamics with a good accuracy.