Title :
Application of radial basis function networks to model electric arc furnaces
Author :
Sadeghian, A.R. ; Lavers, J.D.
Author_Institution :
Dept. of Electr. & Comput. Eng., Toronto Univ., Ont., Canada
Abstract :
This paper presents neural network-based models for electric arc furnaces. The primary objectives are to investigate and justify the application of fast learning neural network methodologies to model electric furnaces, and to compare the results with the existing recorded data. In contrast with conventional techniques where the model is based on a group of mathematically derived and explicit equations, the neural network-based model is inferred using the data obtained from the electric arc furnace and the trained neural network. Hence, the number of oversimplified assumptions used to model the arc is kept to a minimum. It is shown that the trained neural networks successfully approximate the nonlinear behavior of the arc furnace and provides a reasonably accurate model for a device, namely, the electric arc furnace, with ill-defined mathematical model, nonlinear characteristics and time-variant parameters
Keywords :
arc furnaces; identification; learning (artificial intelligence); radial basis function networks; electric arc furnaces; identification; learning; mathematical model; radial basis function neural networks; Application software; Artificial neural networks; Furnaces; Mathematical model; Neural networks; Nonlinear equations; Power system harmonics; Radial basis function networks; Steel; Voltage fluctuations;
Conference_Titel :
Neural Networks, 1999. IJCNN '99. International Joint Conference on
Conference_Location :
Washington, DC
Print_ISBN :
0-7803-5529-6
DOI :
10.1109/IJCNN.1999.830798