Abstract
An electric arc furnace exhibits highly nonlinear characteristics and can cause power quality issues. Its performance is a vital issue for electric utilities. In this paper, a novel multiobjective optimization technique has been proposed for the estimation of the electric arc furnace parameters to predict the voltage and current waveforms and evaluate the performance of the furnace under various conditions. The multiobjective optimization technique minimizes the extinction voltage error and arc resistance error simultaneously to enhance the performance of the electric arc furnace model. The proposed optimization model employs a genetic algorithm and the stochastic variation of the arc length to estimate the parameters of a nonlinear time-variant resistance model of the electric arc furnace by using actual current and voltage data. These data have been obtained from a steel plant in Saudi Arabia. Optimization has been carried out in MATLAB/SIMULINK environment and the model of the electric arc furnace with the estimated parameters has been developed in PSCAD/EMTDC. The results obtained from the PSCAD model have been validated with actual data from the steel industry. The validation shows that the proposed method is efficient and accurate in the estimation of the electric arc furnace parameters.