A hybrid neuro-wavelet approach to electric arc furnace modeling.
thesisposted on 22.05.2021, 10:38 by Shadan Ghaffaripour
This thesis proposes a hybrid neuro-wavelet based approach for modeling the dynamic voltage-current characteristics in electrical arc furnaces. This method uses the data obtained from an operational electrical arc furnace exclusively to describe the underlying process, and unlike conventional mathematical techniques it does not rely on presumed model structures or simplified assumptions. A comparison between the results that proceeded from the proposed method and the actual measurements has been made. The proposed method is demonstrated to be capable of modeling the EAF's dynamic voltage-current behaviour accurately.