A novel hybrid intelligent model for molten iron temperature forecasting based on machine learning

Wei Xu; Jingjing Liu; Jinman Li; Hua Wang; Qingtai Xiao; Wei Xu; Jingjing Liu; Jinman Li; Hua Wang; Qingtai Xiao

doi:10.3934/math.2024061

AIMS Mathematics

2024, Volume 9, Issue 1: 1227-1247. doi: 10.3934/math.2024061

Previous Article Next Article

Research article Special Issues

A novel hybrid intelligent model for molten iron temperature forecasting based on machine learning

1.
State Key Laboratory of Complex Nonferrous Metal Resources Clean Utilization, Kunming University of Science and Technology, Kunming 650093, China
2.
Faculty of Metallurgical and Energy Engineering, Kunming University of Science and Technology, Kunming 650093, China
3.
School of Mathematics, Science and Engineering, University of the Incarnate Word, San Antonio, Texas 78209, United States
4.
Department of Quality Management, Inspection and Testing, Yibin University, Yibin, 644000, China

Received: 31 October 2023 Revised: 27 November 2023 Accepted: 28 November 2023 Published: 07 December 2023
MSC : 62P30

To address the challenges of low accuracy and poor robustness of traditional single prediction models for blast furnace molten iron temperature, a hybrid model that integrates the improved complete ensemble empirical mode decomposition with adaptive noise, kernel principal component analysis, support vector regression and radial basis functional neural network is proposed for precise and stable iron temperature prediction. First, the complete ensemble empirical mode decomposition is employed to decompose the time series of iron temperature, yielding several intrinsic mode functions. Second, kernel principal component analysis is used to reduce the dimensionality of the multi-dimensional key variables from the steel production process, extracting the major features of these variables. Then, in conjunction with the K-means algorithm, support vector regression is utilized to predict the first column of the decomposed sequence, which contains the most informative content, evaluated using the Pearson correlation coefficient method and permutation entropy calculation. Finally, radial basis function neural network is applied to predict the remaining time series of iron temperature, resulting in the cumulative prediction. Results demonstrate that compared to traditional single models, the mean absolute percentage error is reduced by 54.55%, and the root mean square error is improved by 49.40%. This novel model provides a better understanding of the dynamic temperature variations in iron, and achieves a hit rate of 94.12% within a range of ±5℃. Consequently, this work offers theoretical support for real-time control of blast furnace molten iron temperature and holds practical significance for ensuring the stability of blast furnace smelting and implementing intelligent metallurgical processes.
- molten iron temperature,
- intelligent prediction,
- K-means,
- empirical mode decomposition,
- machine learning
Citation: Wei Xu, Jingjing Liu, Jinman Li, Hua Wang, Qingtai Xiao. A novel hybrid intelligent model for molten iron temperature forecasting based on machine learning[J]. AIMS Mathematics, 2024, 9(1): 1227-1247. doi: 10.3934/math.2024061

Related Papers:

Abstract

To address the challenges of low accuracy and poor robustness of traditional single prediction models for blast furnace molten iron temperature, a hybrid model that integrates the improved complete ensemble empirical mode decomposition with adaptive noise, kernel principal component analysis, support vector regression and radial basis functional neural network is proposed for precise and stable iron temperature prediction. First, the complete ensemble empirical mode decomposition is employed to decompose the time series of iron temperature, yielding several intrinsic mode functions. Second, kernel principal component analysis is used to reduce the dimensionality of the multi-dimensional key variables from the steel production process, extracting the major features of these variables. Then, in conjunction with the K-means algorithm, support vector regression is utilized to predict the first column of the decomposed sequence, which contains the most informative content, evaluated using the Pearson correlation coefficient method and permutation entropy calculation. Finally, radial basis function neural network is applied to predict the remaining time series of iron temperature, resulting in the cumulative prediction. Results demonstrate that compared to traditional single models, the mean absolute percentage error is reduced by 54.55%, and the root mean square error is improved by 49.40%. This novel model provides a better understanding of the dynamic temperature variations in iron, and achieves a hit rate of 94.12% within a range of ±5℃. Consequently, this work offers theoretical support for real-time control of blast furnace molten iron temperature and holds practical significance for ensuring the stability of blast furnace smelting and implementing intelligent metallurgical processes.

References

[1]	S. Liu, Q. Xiao, An empirical analysis on spatial correlation investigation of industrial carbon emissions using SNA-ICE model, Energy, 224 (2021), 120183. https://doi.org/10.1016/j.energy.2021.120183 doi: 10.1016/j.energy.2021.120183
[2]	K. He, L. Wang, A review of energy use and energy-efficient technologies for the iron and steel industry, Renew. Sust. Energy Rev., 70 (2017), 1022–1039. https:/doi.org/10.1016/j.rser.2016.12.007 doi: 10.1016/j.rser.2016.12.007
[3]	M. A. Quader, S. Ahmed, R. A. R. Ghazilla, S. Ahmed, M. Dahari, A comprehensive review on energy efficient CO₂ breakthrough technologies for sustainable green iron and steel manufacturing, Renew. Sust. Energy Rev., 50 (2015), 594–614. https:/doi.org/10.1016/j.rser.2015.05.026 doi: 10.1016/j.rser.2015.05.026
[4]	M. Smith, Blast furnace ironmaking: view on future developments, Ironmak. Steelmak., 42 (2015), 734–742. https:/doi.org/10.1179/0301923315z.000000000422 doi: 10.1179/0301923315z.000000000422
[5]	Y. Yang, S. Zhang, Y. Yin, A modified ELM algorithm for the prediction of silicon content in hot metal, Neural Comput. Appl., 27 (2014), 241–247. https:/doi.org/10.1007/s00521-014-1775-x doi: 10.1007/s00521-014-1775-x
[6]	J. Li, X. Wei, H. Chen, Y. Yang, L. Min, Double-hyperplane fuzzy classifier design for tendency prediction of silicon content in molten iron, Fuzzy Set Syst., 426 (2022), 163–175. https:/doi.org/10.1016/j.fss.2021.05.002 doi: 10.1016/j.fss.2021.05.002
[7]	R. D. Martín, F. Obeso, J. Mochón, R. Barea, J. Jiménez, Hot metal temperature prediction in blast furnace using advanced model based on fuzzy logic tools, Ironmak. Steelmak., 34 (2013), 241–247. https://doi.org/10.1179/174328107x155358 doi: 10.1179/174328107x155358
[8]	C. S. Tsao, R. H. Day, A process analysis model of the U.S. steel industry, Manage. Sci., 17 (1971), B569–B704. https:/doi.org/10.1287/mnsc.17.10.B588 doi: 10.1287/mnsc.17.10.B588
[9]	Y. Dai, J. Li, C. Shi, W. Yan, Dephosphorization of high silicon hot metal based on double slag converter steelmaking technology, Ironmak. Steelmak., 48 (2020), 447–456. https:/10.1080/03019233.2020.1807288 doi: 10.1080/03019233.2020.1807288
[10]	J. Y. Liu, W. Zhang, Blast furnace temperature prediction based on RBF neural network and genetic algorithm, Electron. Meas. Technol., 41 (2018), 42–45. https:/10.3390/lubricants9090086 doi: 10.3390/lubricants9090086
[11]	V. R. Radhakrishnan, K. M. Ram, Mathematical model for predictive control of the bell-less top charging system of a blast furnace, J. Process. Control, 11 (2001), 565–586. https://doi.org/10.1016/s0959-1524(00)00026-3 doi: 10.1016/s0959-1524(00)00026-3
[12]	M. Geerdes, R. Chaigneau, O. Lingiardi, R. Molenaar, R. van Opbergen, Y. Sha, et al., Modern blast furnace ironmaking, IOS Press eBooks, 2020. https://doi.org/10.3233/stal9781643681238
[13]	S. Amano, T. Takarabe, T. Nakamori, H. Oda, M. Taira, S. Watanabe, et al., Expert system for blast furnace operation at Kimitsu works, ISIJ Int., 30 (1990), 105–110. https://doi.org/10.2355/isijinternational.30.105 doi: 10.2355/isijinternational.30.105
[14]	T. Yang, S. Yang, G. Zuo, H. Wei, J. Xu, Y. Zhou, An expert system for abnormal status diagnosis and operation guide of a blast furnace, IFAC Proc. Vol., 25 (1992), 59–63. https://doi.org/10.1016/S1474-6670(17)49899-5 doi: 10.1016/S1474-6670(17)49899-5
[15]	E. Lughofer, R. A. Pollak, C. Feilmayr, M. Schatzl, S. Saminger-Platz, Prediction and explanation models for hot metal temperature, silicon concentration, and cooling capacity in ironmaking blast furnaces, Steel Res. Int., 92 (2021), 2100078. https://doi.org/10.1002/srin.202100078 doi: 10.1002/srin.202100078
[16]	J. Jimenez, J. Mochon, J. S. de Ayala, F. Obeso, Blast furnace hot metal temperature prediction through neural networks-based models, ISIJ Int., 44 (2004), 573–580. https://doi.org/10.2355/isijinternational.44.573 doi: 10.2355/isijinternational.44.573
[17]	W. Chen, F. Kong, B. Wang, Y. H. Li, Application of grey relational analysis and extreme learning machine method for predicting silicon content of molten iron in blast furnace, Ironmak. Steelmak., 46 (2018), 974–979. https://doi.org/10.1080/03019233.2018.1470146 doi: 10.1080/03019233.2018.1470146
[18]	J. Song, X. Xing, Z. Pang, M. Lv, Prediction of silicon content in the hot metal of a blast furnace based on FPA-BP model, Metals, 13 (2023), 918. https://doi.org/10.3390/met13050918 doi: 10.3390/met13050918
[19]	W. Liang, G. Wang, X. Ning, J. Zhang, Y. Li, C. Jiang, et al., Application of BP neural network to the prediction of coal ash melting characteristic temperature, Fuel, 260 (2020), 116324. https://doi.org/10.1016/j.fuel.2019.116324 doi: 10.1016/j.fuel.2019.116324
[20]	X. Su, S. Zhang, Y. Yin, W. Xiao, Prediction model of hot metal temperature for blast furnace based on improved multi-layer extreme learning machine, Int. J. Mach. Learn. Cyb., 10 (2019), 2739–2752. https://doi.org/10.1007/s13042-018-0897-3 doi: 10.1007/s13042-018-0897-3
[21]	H. Zhang, Y. Yin, S. Zhang, An improved ELM algorithm for the measurement of hot metal temperature in blast furnace, Neurocomputing, 174 (2015), 232–237. https://doi.org/10.1016/j.neucom.2015.04.106 doi: 10.1016/j.neucom.2015.04.106
[22]	X. Huang, H. Chen, X. Ling, L. Liu, T. Huhe, Investigation of heat and mass transfer and gas-liquid thermodynamic process paths in a humidifier, Energy, 261 (2022), 125156. https://doi.org/10.1016/j.energy.2022.125156 doi: 10.1016/j.energy.2022.125156
[23]	M. M. Li, B. Verma, Nonlinear curve fitting to stopping power data using RBF neural networks, Expert Syst. Appl., 45 (2016), 161–171. https://doi.org/10.1016/j.eswa.2015.09.033 doi: 10.1016/j.eswa.2015.09.033
[24]	K. Yang, J. Li, M. Wang, H. Wang, Q. Xiao, Identifying flow patterns in a narrow channel via feature extraction of conductivity measurements with a support vector machine, Sensors, 23 (2023), 1907. https://doi.org/10.3390/s23041907 doi: 10.3390/s23041907
[25]	P. Zhou, D. Guo, H. Wang, T. Chai, Data-driven robust M-LS-SVR-based NARX modeling for estimation and control of molten iron quality indices in blast furnace ironmaking, IEEE Trans. Neural Networks Learn. Syst., 29 (2018), 4007–4021. https://doi.org/10.1109/tnnls.2017.2749412 doi: 10.1109/tnnls.2017.2749412
[26]	P. Zhou, D. Guo, T. Chai, Data-driven predictive control of molten iron quality in blast furnace ironmaking using multi-output LS-SVR based inverse system identification, Neurocomputing, 308 (2018), 101–110. https://doi.org/10.1016/j.neucom.2018.04.060 doi: 10.1016/j.neucom.2018.04.060
[27]	Y. Liao, Y. Wang, M. Li, Q. Xiao, H. Wang, Prediction for the reduction smelting temperature based on CEEMDAN-RVM-EC, Control Theory Appl., 39 (2022), 2177–2184.
[28]	Y. Wang, P. Yang, S. Zhao, J. Chevallier, Q. Xiao, A hybrid intelligent framework for forecasting short-term hourly wind speed based on machine learning, Expert. Syst. Appl., 213 (2023), 119223. https://doi.org/10.1016/j.eswa.2022.119223 doi: 10.1016/j.eswa.2022.119223
[29]	K. Yang, Y. Wang, M. Li, X. Li, H. Wang, Q. Xiao, Modeling topological nature of gas-liquid mixing process inside rectangular channel using RBF combined with CEEMDAN-VMD, Chem. Eng. Sci., 267 (2023), 118353. https://doi.org/10.1016/j.ces.2022.118353 doi: 10.1016/j.ces.2022.118353
[30]	Y. Wang, P. Yang, Z. Song, J. Chevallier, Q. Xiao, Intelligent prediction of annual CO₂ emissions under data decomposition mode, Comput. Econ., 2023. https://doi.org/10.1007/s10614-023-10357-8 doi: 10.1007/s10614-023-10357-8
[31]	Z. Cui, A. Yang, L. Wang, Y. Han, Dynamic prediction model of silicon content in molten iron based on comprehensive characterization of furnace temperature, Metals, 12 (2022), 1403. https://doi.org/10.3390/met12091403 doi: 10.3390/met12091403
[32]	D. O. L. Fontes, L. G. S. Vasconcelos, R. P. Brito, Blast furnace hot metal temperature and silicon content prediction using soft sensor based on fuzzy C-means and exogenous nonlinear autoregressive models, Comput. Chem. Eng., 141 (2020), 107028. https://doi.org/10.1016/j.compchemeng.2020.107028 doi: 10.1016/j.compchemeng.2020.107028
[33]	J. Zhao, X. Li, L. Song, K. Wang, Q. Lyu, E. Liu, Prediction of hot metal temperature based on data mining, High Temp. Mat. Process., 40 (2021), 87–98. https://doi.org/10.1515/htmp-2021-0020 doi: 10.1515/htmp-2021-0020
[34]	L. J. Cao, K. S. Chua, W. Chong, H. P. Lee, Q. M. Gu, A comparison of PCA, KPCA and ICA for dimensionality reduction in support vector machine, Neurocomputing, 55 (2003), 321–336. https://doi.org/10.1016/s0925-2312(03)00433-8 doi: 10.1016/s0925-2312(03)00433-8
[35]	K. Yang, H. Wang, H. Wang, M. Li, Q. Xiao, Topological approach for the measurement of mixing state quality in a vertical rectangular channel, Int. J. Multiphas. Flow, 163 (2023), 104431. https://doi.org/10.1016/j.ijmultiphaseflow.2023.104431 doi: 10.1016/j.ijmultiphaseflow.2023.104431

Reader Comments

Your name:*

Email:*
© 2024 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)