Blast Furnace System Modeling by Multivariate Phase Space Reconstruction and Neural Networks

Abstract

Time series data collected from a medium‐size blast furnace (BF) is analyzed using the phase space reconstruction. To achieve better reconstruction, multivariate correlation analysis is first applied to screen out correlated variables, which shows that three important variables, i.e., silicon content in hot metal ([Si]), permeability index (FF), and coal injection (PM), are most appropriate for multivariate reconstruction. The time delay and embedding dimension are determined via the autocorrelation function and false nearest neighbor method. With the reconstructed time series, the neural networks model is applied to construct the predictive model for silicon content in hot metal. The simulation shows that the models based on multivariate reconstruction give better predictions than those obtained by univariate reconstruction. Moreover, it reveals that multivariate reconstruction can greatly mitigate the drawbacks caused by insufficiency of data.

Ensemble Non-Gaussian Local Regression for Industrial Silicon Content Prediction

“…And the data in different operating areas are distributed irregularly. To evaluate the prediction performance, three indices of the root-mean-square error (RMSE), relative RMSE (simply noted as RE), and the hit rate (HR) [19][20][21][22][23][24][25][26] are utilized and defined below, respectively. (11) where y q and ŷ q are the actual value and the predicted value, respectively.…”

Adaptive Weighting Just-in-Time-Learning Quality Prediction Model for an Industrial Blast Furnace

“…40) Additionally, the hit rate (HR) index is often adopted in industrial blast furnace ironmaking processes. [21][22][23][24][25][26][27][28] Three indices of RMSE, RE, and HR are defined, respectively. (18) where ŷ q and y q are the predicted value and the actual value, respectively.…”

“…To online predict the silicon content, various data-driven soft sensor modeling approaches, including various neural networks, [7][8][9][10][11][12][13][14] partial least squares, 14,15) fuzzy inference systems, 16) nonlinear time series analysis, [17][18][19][20] subspace identification, 21) support vector regression (SVR) and least squares SVR (LSSVR), [22][23][24] and others [25][26][27][28][29] have been investigated. A recent overview of black-box models for short-term silicon content prediction in blast furnaces can be referred to.…”

Identification of Extreme Temperature Fluctuation in Blast Furnace Based on Fractal Time Series Analysis

“…Thus, the analysis of silicon content will provide the evidence for the temperature control of blast furnace. The previous research had proved that during the iron making process, the silicon content series is non-normal, non-linear and long-term negative correlated, based on these features, C. H. Gao and his team applied a lot of models to predict the silicon content during the ironmaking process, such as: data-driven model based on Volterra series [5], data-driven time discrete models [6], multivariate phase space reconstruction and neural networks model [8], and modelling of the thermal state change of blast furnace hearth with support vector machines [10]. A. Nurkkala, F. Pettersson, and H. Saxén estimated the analyse model based on the non-linear feature [7], while S. Ueda, S. Natsui, and H. Nogami reviewed all the possible mathematical models that may provide help in this research [9].…”