Classification of blast furnace internal state based on FLS and its application in furnace temperature prediction

Abstract

The real-time and accurate prediction of the molten iron silicon content of the blast furnace plays an important role in regulating the temperature of the blast furnace and stabilizing the furnace condition. When the time is large, the accuracy and credibility of the forecast results decrease rapidly, which is not conducive to on-site operators to carry out production operations according to the forecast results. To this end, this paper adds a state variable to each piece of data through the flexible least square parameter estimation method, and selects the training set in a state similar to the test sample. This makes the selection of training data more accurate and reliable. Application examples show that the method proposed in this paper improves the accuracy of silicon content prediction results and has good guiding significance for actual production operations.

Machine learning based knowledge discovery and modeling of silicon content of molten iron from a blast furnace

“…In the case of physics-based models (Abhale et al, 2022(Abhale et al, , 2020Austin et al, 1997;Dong et al, 2007;Roeplal et al, 2023;Yu and Shen, 2022), mass, momentum and enthalpy balance equations are solved for the furnace to predict HMSi and other blast furnace performance indicators such as hot metal temperature, furnace permeability, fuel rate and productivity. In the case of data-driven models (Bhattacharya, 2005;Chuanhou Gao et al, 2011;Diniz et al, 2021;Gao et al, 2021;Gaopeng, 2011;Gao-peng et al, 2021bGao-peng et al, , 2021aJian et al, 2015;Li et al, 2018Li et al, , 2017Li et al, , 2013Liu et al, 2007;Nurkkala et al, 2011;Saxén and Pettersson, 2007;Saxén et al, 2016;Shi-hua and Jiu-sun, 2007;Tang et al, 2009;Wang, 2018;Wang et al, 2019Wang et al, , 2015Wang et al, , 2022Wang and Liu, 2011;Zeng et al, 2008;Zhao et al, 2020), an empirical relationship between silicon content and various raw material and process parameters is established using historical operations data of the furnace through statistical, machine learning and deep learning techniques. Physics-based models are often computationally intensive and typically not suitable for real-time deployment.…”

“…Other techniques used for building prediction models for HMSi included partial least squares (Bhattacharya, 2005;Wang et al, 2019), Bayesian networks (Li et al, 2018;Liu et al, 2007), linear models (Gao-peng et al, 2021a;Saxén et al, 2016) and autoregression (Chuanhou Gao et al, 2011;Gao et al, 2021;Shi-hua and Jiu-sun, 2007;Wang, 2018;Zeng et al, 2008). Bhattacharya (2005) built a partial least squares model for predicting hourly HMSi using 120 process variables from an industrial blast furnace.…”