Scholarlink

<p>Copper matte is the indispensable intermediate product of matte smelting process for further extracting and refining copper. The antimony (Sb) content in copper matte has a serious impact on lining life of refining furnace, cathode copper quality, and precious metal yield in the subsequent process. To improve the prediction accuracy of Sb content in copper matte during the oxygen-enriched bottom-blown continuous smelting process, this paper proposes a novel prediction model based on factor analysis, cubic chaotic mapping, and an inertial weight whale optimization algorithm integrated with backpropagation neural network (FA-CIWOA-BP). Grey Relational Analysis (GRA) and FA are employed to screen and reduce the dimensionality of characteristic parameters. Subsequently, the collected data are trained, learned, and verified using the CIWOA-BP optimization algorithm. The results demonstrate that the model exhibits a coefficient of determination (<i>R</i><sup>2</sup>) of 0.948, with the mean root mean square error (RMSE), mean absolute error (MAE), and the maximum absolute error of 0.00222, 0.00154, and 0.0057, respectively, which indicate that the proposed prediction model possesses reasonable generalization ability, robustness, and accuracy. Consequently, the optimized CIWOA-BP model can provide an effective approach for predicting the antimony content in copper matte, thereby offering a guideline for optimizing mixing raw materials and process parameter control in practical matte smelting production.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Prediction Optimization of Sb Content in Copper Matte Using Hybrid Feature Selection and ClWOA-BP Neural Network

  • 期刊论文

  • 发表日期：

    2025年8月8日

查看全文
  • Jianjun Yang,
  • Yaru Cui,
  • Guohua Wang,
  • Xihong He,
  • Linbo Li,
  • Ziyang Liu,
  • Junfang Nan,
  • Yexin Zhang

摘要

Copper matte is the indispensable intermediate product of matte smelting process for further extracting and refining copper. The antimony (Sb) content in copper matte has a serious impact on lining life of refining furnace, cathode copper quality, and precious metal yield in the subsequent process. To improve the prediction accuracy of Sb content in copper matte during the oxygen-enriched bottom-blown continuous smelting process, this paper proposes a novel prediction model based on factor analysis, cubic chaotic mapping, and an inertial weight whale optimization algorithm integrated with backpropagation neural network (FA-CIWOA-BP). Grey Relational Analysis (GRA) and FA are employed to screen and reduce the dimensionality of characteristic parameters. Subsequently, the collected data are trained, learned, and verified using the CIWOA-BP optimization algorithm. The results demonstrate that the model exhibits a coefficient of determination (R2) of 0.948, with the mean root mean square error (RMSE), mean absolute error (MAE), and the maximum absolute error of 0.00222, 0.00154, and 0.0057, respectively, which indicate that the proposed prediction model possesses reasonable generalization ability, robustness, and accuracy. Consequently, the optimized CIWOA-BP model can provide an effective approach for predicting the antimony content in copper matte, thereby offering a guideline for optimizing mixing raw materials and process parameter control in practical matte smelting production.