The research of end-point prediction based on combination model of electric arc furnace

Abstract

Because the end-point parameters of an electric arc furnace (EAF) are affected by both quantitative factors and non-quantitative factors, we combine Nonlinear Gray Bernoulli-Markov model with Support Vector Machine (SVM) to produce a Nonlinear Gray Bernoulli-Markov-SVM prediction model for estimating the end-point parameter values of an EAF. The effects from the non-quantitative factors on the prediction values of end-point parameters are reflected by Nonlinear Gray Bernoulli-Markov model; while the effects from the quantitative inputs are reflected by the SVM. The Nonlinear Gray Bernoulli model that reflects non-quantitative factors is established firstly, and then, its prediction values are revised by the Markov chain. Because the effect from the quantitative inputs can not be reflected by Nonlinear Gray Bernoulli-Markov model, the Nonlinear Gray Bernoulli-Markov model is certainly not free from prediction errors from the quantitative inputs. These prediction errors are compensated by the SVM model. Further, combining Differential Evolution (DE) with Particle Swarm Optimization (PSO), a Differential Evolution Particle Swarm Optimization (DEPSO) is proposed. The parameter of Nonlinear Gray Bernoulli Model is optimized by the PSO algorithm, while that of SVM are optimized by the DEPSO algorithm. The final prediction values of the end-point parameters in EAF are thus obtained. Simultaneously, a rolling forecasting mechanism is introduced for improving the prediction precision.