留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码
Volume 31 Issue 12
Dec.  2024
Turn off MathJax
目录

图(10)  / 表(5)

数据统计

分享

计量
  • 文章访问数:  307
  • HTML全文浏览量:  102
  • PDF下载量:  92
  • 被引次数: 0
文章导航 >  矿物冶金与材料学报（英文）  > 2024  >  31(12): 2657-2669
Zicheng Xin, Jiangshan Zhang, Kaixiang Peng, Junguo Zhang, Chunhui Zhang, Jun Wu, Bo Zhang,  and Qing Liu, Explainable machine learning model for predicting molten steel temperature in the LF refining process, Int. J. Miner. Metall. Mater., 31(2024), No. 12, pp. 2657-2669. https://doi.org/10.1007/s12613-024-2950-4
Cite this article as:
Zicheng Xin, Jiangshan Zhang, Kaixiang Peng, Junguo Zhang, Chunhui Zhang, Jun Wu, Bo Zhang,  and Qing Liu, Explainable machine learning model for predicting molten steel temperature in the LF refining process, Int. J. Miner. Metall. Mater., 31(2024), No. 12, pp. 2657-2669. https://doi.org/10.1007/s12613-024-2950-4
shu
引用本文 PDF XML SpringerLink
研究论文

基于可解释性机器学习的LF精炼钢水温度预测模型

  • 1)

    北京科技大学绿色低碳钢铁冶金全国重点实验室, 北京 100083, 中国

  • 2)

    北京科技大学自动化学院, 北京 100083, 中国

  • 3)

    河钢股份有限公司唐山分公司, 唐山 063000, 中国

  • 4)

    宝钢集团新疆八一钢铁有限公司, 乌鲁木齐, 830022, 中国

  • 5)

    吕梁建龙实业有限公司, 吕梁, 032100, 中国



  • 通讯作者:

    张江山    E-mail: qliu@ustb.edu.cn

    刘青    E-mail: zjsustb@163.com

文章亮点

  • (1) 构建了灰狼算法优化轻量梯度提升机的LF精炼钢水温度预测模型
  • (2) 利用树形结构可视化揭示了LF精炼钢水温度预测模型的学习/决策过程
  • (3) 基于SHAP分析明确了不同输入变量对钢水温度的影响程度
  • 钢包炉（LF）精炼过程中，钢水温度的精准预测对钢水质量和炼钢成本的控制具有重要影响。当前针对钢水温度预测模型开展了大量研究工作，然而，大多数研究学者主要关注预测模型精度的提高，忽视了预测模型的可解释性。基于此，本文首先采用极端梯度提升树（XGBoost）和轻量梯度提升机（LGBM），融合贝叶斯优化和灰狼优化（GWO）来建立LF精炼钢水温度预测模型。然后，通过比较不同模型的性能评价指标，得出了运用不同超参数优化方法获得的最优XGBoost模型和LGBM模型。结果表明，在钢水温度预测中GWO-LGBM模型性能优于其他模型，且在±5°C的误差范围内，模型预测精度达到89.35%。最后，利用树形结构可视化和SHAP分析，揭示了模型的学习/决策过程，明确了不同输入变量对钢水温度的影响程度。该研究有利于操作人员更好地了解输入变量对预测结果的影响规律，从而实现模型的可靠应用和调试，指导操作人员对模型参数进行调整。
    • Research Article

    Explainable machine learning model for predicting molten steel temperature in the LF refining process

    + Author Affiliations
      Abstract
    • Accurate prediction of molten steel temperature in the ladle furnace (LF) refining process has an important influence on the quality of molten steel and the control of steelmaking cost. Extensive research on establishing models to predict molten steel temperature has been conducted. However, most researchers focus solely on improving the accuracy of the model, neglecting its explainability. The present study aims to develop a high-precision and explainable model with improved reliability and transparency. The eXtreme gradient boosting (XGBoost) and light gradient boosting machine (LGBM) were utilized, along with bayesian optimization and grey wolf optimization (GWO), to establish the prediction model. Different performance evaluation metrics and graphical representations were applied to compare the optimal XGBoost and LGBM models obtained through varying hyperparameter optimization methods with the other models. The findings indicated that the GWO-LGBM model outperformed other methods in predicting molten steel temperature, with a high prediction accuracy of 89.35% within the error range of ±5°C. The model’s learning/decision process was revealed, and the influence degree of different variables on the molten steel temperature was clarified using the tree structure visualization and SHapley Additive exPlanations (SHAP) analysis. Consequently, the explainability of the optimal GWO-LGBM model was enhanced, providing reliable support for prediction results.
      • FullText(HTML)
    • loading
      • Supplements(1)
    • Supplementary Information-s12613-024-2950-4.docx
      • References(46)
    • [1]
      R.Y. Yin, A discussion on “smart” steel plant—View from physical system side, Iron Steel, 52(2017), No. 6, p. 1.
      [2]
      Z.C. Xin, J.S. Zhang, Y. Jin, J. Zheng, and Q. Liu, Predicting the alloying element yield in a ladle furnace using principal component analysis and deep neural network, Int. J. Miner. Metall. Mater., 30(2023), No. 2, p. 335. doi: 10.1007/s12613-021-2409-9
      [3]
      Q. Liu, X. Shao, J.P. Yang, and J.S. Zhang, Multiscale modeling and collaborative manufacturing for steelmaking plants, Chin. J. Eng., 43(2021), No. 12, p. 1698.
      [4]
      Z.C. Xin, J.S. Zhang, K.X. Peng, J.G. Zhang, C.H. Zhang, and Q. Liu, Modeling of LF refining process: A review, J. Iron Steel Res. Int., 31(2024), No. 2, p. 289. doi: 10.1007/s42243-023-01100-6
      [5]
      N.L. Samways and T.E. Dancy, Factors affecting temperature drop between tapping and teeming, JOM, 12(1960), No. 4, p. 331. doi: 10.1007/BF03377980
      [6]
      A. Zimmer, Á.N.C. Lima, R.M. Trommer, S.R. Bragança, and C.P. Bergmann, Heat transfer in steelmaking ladle, J. Iron Steel Res. Int., 15(2008), No. 3, p. 11. doi: 10.1016/S1006-706X(08)60117-X
      [7]
      Y.J. Wu, Z.H. Jiang, M.F. Jiang, W. Gong, and D.P. Zhan, Temperature prediction model of molten steel in LF, J. Iron Steel Res., 14(2002), p. 9.
      [8]
      P.S. Srinivas, A.K. Kothari, and A. Agrawal, Parameter estimation by inverse solution methodology using genetic algorithms for real time temperature prediction model of ladle furnace, ISIJ Int., 56(2016), No. 6, p. 977. doi: 10.2355/isijinternational.ISIJINT-2015-712
      [9]
      A.N. Wang, H.X. Tian, Z.H. Jiang, D.P. Zhan, X.D. Yin, and Z.G. Ma, Temperature prediction of molten steel for LF based on information fusion technique, J. Iron Steel Res., 17(2005), No. 6, p. 71.
      [10]
      Z.C. Xin, J.S. Zhang, J.G. Zhang, J. Zheng, Y. Jin, and Q. Liu, Predicting temperature of molten steel in LF-refining process using IF–ZCA–DNN model, Metall. Mater. Trans. B, 54(2023), No. 3, p. 1181. doi: 10.1007/s11663-023-02753-0
      [11]
      W. Lv, Z.Z. Mao, P. Yuan, and M.X. Jia, Multi-kernel learnt partial linear regularization network and its application to predict the liquid steel temperature in ladle furnace, Knowl. Based Syst., 36(2012), p. 280. doi: 10.1016/j.knosys.2012.07.012
      [12]
      F. Yuan, A.J. Xu, and M.Q. Gu, Development of an improved CBR model for predicting steel temperature in ladle furnace refining, Int. J. Miner. Metall. Mater., 28(2021), No. 8, p. 1321. doi: 10.1007/s12613-020-2234-6
      [13]
      H.X. Tian, Y.D. Liu, K. Li, R.R. Yang, and B. Meng, A new AdaBoost.IR soft sensor method for robust operation optimization of ladle furnace refining, ISIJ Int., 57(2017), No. 5, p. 841. doi: 10.2355/isijinternational.ISIJINT-2016-371
      [14]
      Z.C. Xin, J.S. Zhang, J. Zheng, Y. Jin, and Q. Liu, A hybrid modeling method based on expert control and deep neural network for temperature prediction of molten steel in LF, ISIJ Int., 62(2022), No. 3, p. 532. doi: 10.2355/isijinternational.ISIJINT-2021-251
      [15]
      Ü. Çamdali and M. Tunc, Energy and exergy analysis of a ladle furnace, Can. Metall. Q., 42(2003), p. 439. doi: 10.1179/cmq.2003.42.4.439
      [16]
      Z.Y. Tao, M.F. Jiang, and C.J. Liu, Prediction of molten steel end point temperature in LF based on modified artificial neural network, Spec. Steel, 27(2006), No. 6, p. 21.
      [17]
      G.Q. Fu, Q. Liu, Z. Wang, et al., Grey box model for predicting the LF end-point temperature of molten steel, J. Univ. Sci. Technol. Beijing, 35(2013), No. 7, p. 948.
      [18]
      L.L. Zou, J.S. Zhang, Y.S. Han, F.Z. Zeng, Q.H. Li, and Q. Liu, Internal crack prediction of continuous casting billet based on principal component analysis and deep neural network, Metals, 11(2021), No. 12, art. No. 1976. doi: 10.3390/met11121976
      [19]
      G. Xu, M. Li, Z.M. Lv, and J.W. Xu, Online intelligent product quality monitoring method based on machine learning, Chin. J. Eng., 44(2022), No. 4, p. 730.
      [20]
      X. Shao, Q. Liu, Z.C. Xin, J.S. Zhang, T. Zhou, and S.S. Li, Hybrid model for BOF oxygen blowing time prediction based on oxygen balance mechanism and deep neural network, Int. J. Miner. Metall. Mater., 31(2024), No. 1, p. 106. doi: 10.1007/s12613-023-2670-1
      [21]
      Y. Niu, Z.Q. Hong, Y.Q. Wang, and Y.C. Zhu, Machine learning-based beta transus temperature prediction for titanium alloys, J. Mater. Res. Technol., 23(2023), p. 515. doi: 10.1016/j.jmrt.2023.01.019
      [22]
      T. Shafighfard, F. Bagherzadeh, R.A. Rizi, and D.Y. Yoo, Data-driven compressive strength prediction of steel fiber reinforced concrete (SFRC) subjected to elevated temperatures using stacked machine learning algorithms, J. Mater. Res. Technol., 21(2022), p. 3777. doi: 10.1016/j.jmrt.2022.10.153
      [23]
      P. Mahajan, S. Uddin, F. Hajati, and M.A. Moni, Ensemble learning for disease prediction: A review, Healthcare, 11(2023), No. 12, art. No. 1808.
      [24]
      H.Y. Li, X.J. Liu, X. Li, X.P. Bu, H.W. Li, and Q. Lyu, Causality analysis and prediction of blast furnace state based on convergence cross mapping, Ironmaking Steelmaking, 49(2022), No. 9, p. 875. doi: 10.1080/03019233.2022.2065447
      [25]
      Y.K. Kong and K. Kurumisawa, Application of machine learning in predicting workability for alkali-activated materials, Case Stud. Constr. Mater., 18(2023), art. No. e02173.
      [26]
      M. Pelikan, D.E. Goldberg, and E. Cantú-Paz, BOA: The Bayesian optimization algorithm, [in] Proceedings of the Genetic and Evolutionary Computation Conference, Orlando, 1999, p. 525.
      [27]
      H. Huang, J.N. Qiu, and K. Riedl, On the global convergence of particle swarm optimization methods, Appl. Math. Optim., 88(2023), No. 2, art. No. 30. doi: 10.1007/s00245-023-09983-3
      [28]
      J.H. Holland, Genetic algorithms, Sci. Am., 267(1992), No. 1, p. 66. doi: 10.1038/scientificamerican0792-66
      [29]
      S. Mirjalili, S.M. Mirjalili, and A. Lewis, Grey wolf optimizer, Adv. Eng. Software, 69(2014), p. 46. doi: 10.1016/j.advengsoft.2013.12.007
      [30]
      R.F. Cui, H. Ma, C.P. Guo, H.Y. Li, and Z. Liu, Prediction of missile aerodynamic data based on gradient boosting under Bayesian hyperparametric optimization, Aeronaut. Sci. Technol., 34(2023), No. 7, p. 22.
      [31]
      B. Gong, Study of PLSR-BP model for stability assessment of loess slope based on particle swarm optimization, Sci. Rep., 11(2021), No. 1, art. No. 17888. doi: 10.1038/s41598-021-97484-0
      [32]
      Y.B. He, H. Guo, B.Q. Zhang, Q. Zhu, H.M. Tang, and Y.H. Li, Prediction method of hot metal silicon content based on improved BP neural network, J. Iron. Steel. Res., 36(2024), No. 3, p. 309.
      [33]
      J. Zhou, Y.L. Zhang, C.Q. Li, et al., Enhancing the performance of tunnel water inflow prediction using Random Forest optimized by Grey Wolf Optimizer, Earth Sci. Inform., 16(2023), No. 3, p. 2405. doi: 10.1007/s12145-023-01042-3
      [34]
      M.T. Ribeiro, S. Singh, and C. Guestrin, “Why should I trust you?”: Explaining the predictions of any classifier, [in] Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Diego, 2016, p. 1135.
      [35]
      S.M. Lundberg and S.I. Lee, A unified approach to interpreting model predictions, [in] Proceedings of the 31st Conference and Workshop on Neural Information Processing Systems, California, 2017, p. 4765.
      [36]
      Y.L. Huang, X.L. Qin, Y.W. Chen, L.G. Zhang, and B. Yi, Interpretability analysis of sepsis prediction model using LIME, J. Comput. Appl., 41(2021), No. S1, p. 332.
      [37]
      N. Bar, T. Korem, O. Weissbrod, et al., A reference map of potential determinants for the human serum metabolome, Nature, 588(2020), No. 7836, p. 135. doi: 10.1038/s41586-020-2896-2
      [38]
      W. Liu, Z.X. Chen, Y. Hu, and J. Zhang, Forecasting pipeline safety and remaining life with machine learning methods and SHAP interaction values, Int. J. Press. Vessels Pip., 205(2023), art. No. 105000. doi: 10.1016/j.ijpvp.2023.105000
      [39]
      A. Homafar, H. Nasiri, and S.C. Chelgani, Modeling coking coal indexes by SHAP-XGBoost: Explainable artificial intelligence method, Fuel Commun., 13(2022), art. No. 100078. doi: 10.1016/j.jfueco.2022.100078
      [40]
      S.L. Ma, Theoretical analysis and calculation of LF fume parameters, Environ. Eng., 25(2007), No. 5, p. 44.
      [41]
      H.X. Tian, Z.Z. Mao, and A.N. Wang, Hybrid modeling for soft sensing of molten steel temperature in LF, J. Iron Steel Res. Int., 16(2009), No. 4, p. 1. doi: 10.1016/S1006-706X(09)60051-0
      [42]
      J. Li, J. Fu, P. Wang, Q.F. Bi, and C.G. Huang, Composition adjustment and temperature prediction of molten steel in LF, J. Iron Steel Res., 11(1999), No. 2, p. 6.
      [43]
      S.M. Malakouti, Babysitting hyperparameter optimization and 10-fold-cross-validation to enhance the performance of ML methods in predicting wind speed and energy generation, Intell. Syst. Appl., 19(2023), art. No. 200248. doi: 10.1016/j.iswa.2023.200248
      [44]
      K.E. Taylor, Summarizing multiple aspects of model performance in a single diagram, J. Geophys. Res., 106(2001), No. D7, p. 7183. doi: 10.1029/2000JD900719
      [45]
      S. Mangalathu, S.H. Hwang, and J.S. Jeon, Failure mode and effects analysis of RC members based on machine-learning-based SHapley Additive exPlanations (SHAP) approach, Eng. Struct., 219(2020), art. No. 110927. doi: 10.1016/j.engstruct.2020.110927
      [46]
      B. Xi, E.M. Li, Y. Fissha, J. Zhou, and P. Segarra, LGBM-based modeling scenarios to compressive strength of recycled aggregate concrete with SHAP analysis, Mech. Adv. Mater. Struct., 31(2023), No. 23, p. 5999. doi: 10.1080/15376494.2023.2224782
      • Relative Articles
      Cited By

    Catalog


    • /

      返回文章
      返回

      版权所有 ©《矿物冶金与材料学报（英文）》编辑部

      地址：北京市海淀区学院路30号邮政编码：100083

      电子邮箱：journal@ustb.edu.cn电话：+86-10-62332875

      本系统由 北京仁和汇智信息技术有限公司 开发    百度统计