This paper is devoted to electric arc furnace (EAF) modeling using a random differential equation based on the power balance approach. It presents a method which enables the estimation of EAF model coefficients with the help of measurement data - voltage and current waveforms recorded during the melting stage of an EAF work cycle. The estimated coefficients have been analyzed with respect to their time variability as well as the probability distributions of their values and increments. Next, the identification of the stochastic processes representing the model coefficients has been carried out. The paper presents solutions validated with statistical tests.
1 October 2016 in Statistics & Probability Letters