A Blast Furnace Prediction Model Combining Neural Network with Partial Least Square Regression

Hao, Xinlong; Shen, Fengman; Du, Gang; Shen, Yansong; Xie, Zhi

doi:10.1002/srin.200506080

Cited by 43 publications

(21 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Meanwhile, for the proposed auto parameter setting strategy defined in Eqs. (14) and (15) (PC auto ), the computation is more efficient because only a generalized eigen problem is solved. As illustrated in Fig.…”

Section: Performance Of Utilizing the Output Informationmentioning

confidence: 99%

“…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.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

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

Chen

Liu

2017

ISIJ Int.

View full text Add to dashboard Cite

Development of accurate soft sensors for online quality prediction (e.g., silicon content) in an industrial blast furnace is a difficult task. A novel just-in-time-learning (JITL) prediction approach using adaptive feature-weighting for similar samples is developed. First, a dual-objective joint-optimization framework is proposed to introduce both input and output information into the model. Then, a suitable similarity criterion with feature weighting strategy is formulated, which is not considered in conventional JITL methods. Moreover, the trade-off parameter in the joint-optimization problem can be chosen automatically, without the time-consuming cross-validation procedure. The proposed method is applied to online predict the silicon content in an industrial blast furnace in China. Compared with other JITL-based soft sensors, better prediction performance has been obtained.

show abstract

Section: Performance Of Utilizing the Output Informationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

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

Chen

Liu

2017

ISIJ Int.

View full text Add to dashboard Cite

show abstract

“…(2) Binary coding support vector machines algorithm, [1,5,6] neural network models, [7][8][9][10] and genetic algorithms [11,12] all based on data-mining methods and used to help optimizing control parameters of the BF. (3) Heat and mass balance models.…”

Section: Introductionmentioning

confidence: 99%

Process Simulation and Control Optimization of a Blast Furnace Using Classical Thermodynamics Combined to a Direct Search Algorithm

Harvey

Gheribi

2013

Metall Mater Trans B

View full text Add to dashboard Cite

Several numerical approaches have been proposed in the literature to simulate the behavior of modern blast furnaces: finite volume methods, data-mining models, heat and mass balance models, and classical thermodynamic simulations. Despite this, there is actually no efficient method for evaluating quickly optimal operating parameters of a blast furnace as a function of the iron ore composition, which takes into account all potential chemical reactions that could occur in the system. In the current study, we propose a global simulation strategy of a blast furnace, the 5-unit process simulation. It is based on classical thermodynamic calculations coupled to a direct search algorithm to optimize process parameters. These parameters include the minimum required metallurgical coke consumption as well as the optimal blast chemical composition and the total charge that simultaneously satisfy the overall heat and mass balances of the system. Moreover, a Gibbs free energy function for metallurgical coke is parameterized in the current study and used to fine-tune the simulation of the blast furnace. Optimal operating conditions and predicted output stream properties calculated by the proposed thermodynamic simulation strategy are compared with reference data found in the literature and have proven the validity and high precision of this simulation.

show abstract

“…On the other hand, in an integrated steel works, ironmaking BF is the major energy consumer, accounting for nearly 70% of the whole plant [1][2][3][4].…”

Section: Introductionmentioning

confidence: 99%

Modelling ironmaking blast furnace: Solid flow and thermochemical behaviours

Shen

Guo

et al. 2013

AIP Conference Proceedings

Self Cite

View full text Add to dashboard Cite

Abstract. Ironmaking blast furnace is a counter-, co-, cross-current moving bed reactor, where solid particles are charged at the furnace top forming a downward moving bed while gas are introduced at the lower part of furnace and travels upward through the solid bed of varying porosity, reducing solid ore to liquid iron at the cohesive zone. These three phases interact intensely. In this paper, a three-dimensional mathematical model is developed. The model describes the motion of solid and gas, based on continuum approach, and implements the so-called force balance model for the liquid flow. The model is applied to a blast furnace, where raceway cavity is considered explicitly. The results demonstrate and characterize the key multiphase flow patterns of solid-gas-liquid at different regions inside the blast furnace, in particular solid flow and associated thermochemical behaviours of solid particles. This model offers a costeffective tool to understand and optimize blast furnace operation.

show abstract

A Blast Furnace Prediction Model Combining Neural Network with Partial Least Square Regression

Cited by 43 publications

References 4 publications

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

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

Process Simulation and Control Optimization of a Blast Furnace Using Classical Thermodynamics Combined to a Direct Search Algorithm

Modelling ironmaking blast furnace: Solid flow and thermochemical behaviours

Contact Info

Product

Resources

About