Applying Wavelet Transform and Neural Network for Modeling, Simulation, and Analysis of Dynamic Characteristics of Electric Arc Furnace Loads

• Main Authors: Min-Fu Shih, 施旻甫
• Other Authors: Wen-Kung Chang
• Format: Others
• Language: zh-TW
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/60530138745618126412

碩士 === 國立中正大學 === 電機工程研究所 === 101 === Electric arc furnaces (EAFs) used for steel production are a main source of voltage fluctuations in electrical networks, which may give rise to the flicker effect and causes the problem in the power system. It is necessary to develop a generalized and accurate EAF model for such studies. It is known that artificial neural network is a powerful scheme for function learning and modeling nonlinear loads. This paper proposed a discrete wavelet transform (DWT) and radial basis function neural network (RBFNN) based method for modeling the dynamic voltage current characteristics of the arc electric arc furnace (EAF). In this study, a combination of the DWT and the sequential RBFNN with parameters initialization algorithm is proposed to build the EAF voltage-current characteristics with enhanced look-up table for different operation stages. Simulation results obtained by using the proposed model are compared with different measure data. It shows that the solution procedure accurately models the EAF dynamic voltage-current behavior. Finally, according to the proposed EAF load used MATLAB/SIMULINK for a complete steel plant power system simulation. The proposed method also can be applied to model other highly nonlinear loads and to assess the effectiveness of devices for mitigating the associated power system disturbances or to perform relative penetration studies.