Application of shallow neural networks in electric arc furnace modelling

Abstract

Electric arc furnaces (EAFs) are important appliances in steelmaking industry, but they are characterized by nonlinear, dynamic and stochastic nature. Due to this fact, EAFs can have a negative influence on power systems. Measures for mitigation of such problems can be designed properly only with the knowledge of the load influence on the system. Therefore, it is necessary to have accurate models of EAFs, reflecting the complicated character of such loads. Researchers use different approaches for EAF modelling, such as stochastic processes analysis, differential equation models or neural networks. This paper presents the application of three artificial neural network (ANN) based models in EAF modelling. The goal was to provide ANN models which are simple in structure in comparison to deep learning methods used by other researchers. First two models are built on multilayer perceptron networks and the third applies a nonlinear autoregressive exogenous model with the help of a differential equation transformed into Hammerstein-Wiener model. The paper describes the measurement data, a design for each approach and the results of EAF modelling.