Resumo

MARULANDA DURANGO, Jesser James; ALVAREZ LOPEZ, Mauricio Alexander  e  ALZATE GOMEZ, Alfonso. ESTIMATING ARC FURNACE MODEL PARAMETERSUSING BAYESIAN LINEAR REGRESSION. Cienc. Ing. Neogranad. [online]. 2013, vol.23, n.2, pp.119-136. ISSN 0124-8170.

ABSTRACT In this paper, the authors present the calibration of the parameter of arc furnace that considers the non-linearity and high variability of this type of load. Starting with a nonlinear differential equation that describes the voltage-current characteristic of the arc, an equivalent linear equation that simplifies the estimation of the original model parameters is proposed. Parameter estimation is then accomplished using Bayesian linear regression using measurements taken during the furnance’s most critical operation point. Relationships between the estimated value for the parameters and their uncertainty, in terms of the number of observations included in the model calibration process, are shown. Results obtained through simulation with the estimated parameters are contrasted against real data. A flicker meter, which complies with the IEC standard 61000-4-15, is used for determining the instantaneous flicker level (IFL) due to fluctuations present in the real and simulated current waveforms. Finally, the harmonic content of the real and simulated current waveforms is compared.

Palavras-chave : arc furnace; Chua’s circuit; Bayesian inference.

        · resumo em Espanhol     · texto em Espanhol     · Espanhol ( pdf )