Fractional-Order Model of Electric Arc Furnace

The paper introduces a fractional-order electric arc furnace (EAF) model. It is based on a power balance equation which, as has been shown, can be represented by a Hammerstein-Wiener model. Such a model includes a block described by a linear integer-order differential equation, in this case the order is equal to one, which can be extended to a fractional-order equation. The introduced fractional order is either constant or variable in time. The variable coefficients of the model have been identified by a genetic algorithm The proposed model has been tested using measurement data from two different EAFs and its accuracy has been compared with the classical integer-order model based on the RMS error between the model output and measurement waveforms. The results show that the fractional-order EAF model enables error reduction and consequently reflects more accurately the phenomena taking place in the furnace.