A Sliding‐window Smooth Support Vector Regression Model for Nonlinear Blast Furnace System

Abstract

Blast furnace is one of the most complex industrial reactors and remains some unsolved puzzles, such as blast furnace automation, prediction of the inner thermal state, etc. In this work, a sliding‐window smooth support vector regression model is presented to address the issue of predicting the blast furnace inner thermal state, represented by the silicon content in blast furnace hot metal in the context. Different from the traditional numerical prediction models of silicon, the constructed SW‐SSVR model is devoted to predicting the changing trend of silicon and exhibits good performance with high percentage of successful trend prediction, competitive computational speed and timely online service. Additionally, some sharp fluctuation trend in the silicon test data can also be followed well by the SW‐SSVR model, which is always difficult for traditional data–driven based silicon prediction models. All of these indicate that the SW‐SSVR model is a good candidate to predict the change of blast furnace inner thermal state, and may provide a guide for operators to take proper action on operating blast furnace in advance.

Life cycle inventory processes of the Mittal Steel Poland (MSP) S.A. in Krakow, Poland—blast furnace pig iron production—a case study

“…The properties of BF gas depend on the fuel rate in the furnace, extent of oxygen enrichment of the blast, amount of indirect reduction in the furnace shaft, and other operating conditions. Jian et al (2011) show that BF is a highly complex nonlinear system characterized by high temperature, high pressure, strong noise, and distributed parameters. Generally, the variables related to the raw material and the blast are often taken as input while the variables associated with the hot metal and the in-furnace thermal state are regarded as output (Agarwal et al 2010).…”

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

“…The process input variables correlated with the product quality (i.e., the silicon content) have been selected. 21,22,24) These input variables include the blast volume, the blast temperature, the top pressure, the gas permeability, the top temperature, the ore/coke ratio, and the pulverized coal injection. The sampling time of most of these input variables is 1 minute.…”

“…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.…”

“…[30][31][32] Among them, SVR and LSSVR have shown promising prediction performance, especially when the training data are insufficient. [22][23][24]33) One main disadvantage of most existing data-driven modeling approaches for the silicon content is that a single global model is built. However, it is not enough to describe all the process characteristics only using a single model, [34][35][36][37][38][39][40] especially for some complicated regions with insufficient information.…”