1.承德建龙特殊钢有限公司,承德 067201
2.河北省半钢水冶炼高洁净高品质特殊钢重点实验室,承德 067201
3.河北省发展和改革委员会产业转型升级服务中心,石家庄 050000
王雪原(1976—),男,本科,高级工程师; E-mail : wangxueyuan@ejianlong.com
周春芳(1986—),女,本科,高级工程师; E-mail : zhouchunfang@ejianlong.com;Editorial Office of Special Steel. OA under CC BY-NC-ND 4.0
收稿:2025-05-23,
纸质出版:2026-03-30
移动端阅览
王雪原,张力彬,周春芳.基于UMAP-GWO-DNN的转炉终点磷含量预测模型[J].特殊钢,2026,47(02):41-50.
Wang Xueyuan,Zhang Libin,Zhou Chunfang.Prediction Model of End-point Phosphorus Content in Converter Based on UMAP-GWO-DNN[J].Special Steel,2026,47(02):41-50.
王雪原,张力彬,周春芳.基于UMAP-GWO-DNN的转炉终点磷含量预测模型[J].特殊钢,2026,47(02):41-50. DOI: 10.20057/j.1003-8620.2025-00142.
Wang Xueyuan,Zhang Libin,Zhou Chunfang.Prediction Model of End-point Phosphorus Content in Converter Based on UMAP-GWO-DNN[J].Special Steel,2026,47(02):41-50. DOI: 10.20057/j.1003-8620.2025-00142.
转炉终点磷含量的精准控制是提升钢材质量和冶炼效率的核心环节。本研究针对42CrMo钢种,创新性地融合统一流形逼近与投影(UMAP)、灰狼优化算法(GWO)和深度神经网络(DNN)技术,构建了多模态智能预测模型。通过UMAP算法对高维冶炼参数(如温度、氧枪高度、渣碱度等)进行非线性降维,有效提取关键特征;采用GWO优化DNN的初始权重和超参数,显著提升模型收敛速度与稳定性。实验基于钢厂200炉次实际生产数据,对比BP神经网络、标准DNN及GWO-DNN模型,UMAP-GWO-DNN模型在±0.001%和±0.002%误差区间的命中率分别达到86.7%和95.4%,均方根误差(RMSE)降低23.6%。工业验证表明,该模型使终点磷含量波动标准差减少41%,平均值从0.001 2%稳定至0.000 9%,成功实现"窄窗口"控制目标。本研究为转炉冶炼过程的数字化升级提供了可推广的技术路径。
Precise control of the endpoint phosphorus content in converter steelmaking is a core aspect for enhancing steel quality and smelting efficiency. This study innovatively integrates Uniform Manifold Approximation and Projection (UMAP), Grey Wolf Optimization (GWO), and Deep Neural Network (DNN) technologies to construct a multimodal intelligent prediction model for 42CrMo steel. The UMAP algorithm is employed to perform nonlinear dimensionality reduction on high-dimensional smelting parameters (such as temperature, oxygen lance height, slag basicity, etc.), effectively extracting key features. The GWO algorithm is used to optimize the initial weights and hyperparameters of the DNN, significantly improving the model's convergence speed and stability. The experiments are conducted based on actual production data from 20eats in a steel plant. Compared with the BP neural network, standard DNN, and GWO-DNN models, the UMAP-GWO-DNN model achieves hit rates of 86.7% and 95.4% in error ranges of ±0.001% and ±0.002%, respectively, and the root mean square error (RMSE) is reduced by 23.6%.Industrial validation shows that this model reduces the standard deviation of endpoint phosphorus content fluctuations by 41%, stabilizing the mean value from 0.001 2% to 0.000 9%, successfully achieving the "narrow window" control target. This study provides a scalable technical path for the digital upgrade of converter smelting processes.
Liu X , Bao Y P , Gu C , et al . Distribution and precipitation mechanism of TiN in industrial pure iron bloom [J]. Steel Research International , 2023 , 94 ( 3 ): 2200481 .
Liu X , Wang D Z , Li Z W , et al . Efficient separation of iron elements from steel slag based on magnetic separation process [J]. Journal of Materials Research and Technology , 2023 , 23 : 2362 - 2370 .
Tian B H , Wei G S , Li X , et al . Effect of hot metal charging on economic and environmental indices of electric arc furnace steelmaking in China [J]. Journal of Cleaner Production , 2022 , 379 : 134597 .
Wang Z L , Bao Y P , Gu C . Convolutional neural network-based method for predicting oxygen content at the end point of converter [J]. Steel Research International , 2023 , 94 ( 1 ): 2370011 .
Liu X , Bao Y P , Zhao L H , et al . Establishment and application of steel composition prediction model based on t-distributed stochastic neighbor embedding (t-SNE) dimensionality reduction algorithm [J]. Journal of Sustainable Metallurgy , 2024 , 10 ( 2 ): 509 - 524 .
Xue C , Xu W L , Bao Y P , et al . Prediction of converter tapping weight based on principal component analysis-whale optimization algorithm-backpropagation algorithm neural network model [J]. Steel Research International , 2024 , 95 ( 3 ): 2300410 .
Wang D Z , Gao F , Xing L D , et al . Continuous prediction model of carbon content in 12 converter blowing process [J]. Metals , 2022 , 12 ( 1 ): 151 .
刘玉潇 , 董艳伍 , 姜周华 , 等 . 基于MI和XGBoost算法电渣重熔终点磷含量预报模型 [J]. 特殊钢 , 2025 , 46 ( 1 ): 117 - 125 .
刘增山 , 冯亮花 , 康小兵 . 基于PSO-SVM模型的转炉终点预测 [J]. 特殊钢 , 2024 , 45 ( 3 ): 27 - 32 .
Xin Z C , Zhang J S , Zheng J , et al . A hybrid modeling method based on expert control and deep neural network for temperature prediction of molten steel in LF [J]. ISIJ International , 2022 , 62 ( 3 ): 532 - 541 .
Konovalenko I. , Maruschak P. , Brezinová J. , et al . Steel surface defect classification using deep residual neural network . Metals , 2020 , 10 ( 6 ), 846 .
Song Y J , Li S Z . Gas leak detection in galvanised steel pipe with internal flow noise using convolutional neural network [J]. Process Safety and Environmental Protection , 2021 , 146 : 736 - 744 .
Zhao B , Wu W , Wu W , et al . Study on the occurrence of phosphorus in the slag of hot metal dephosphorization for stainless steel production [J]. Steel Research International , 2020 , 91 ( 7 ): 2000021 .
Ye G F , Yang J , Zhang R H , et al . Behavior of phosphorus enrichment in dephosphorization slag at low temperature and low basicity [J]. International Journal of Minerals, Metallurgy and Materials , 2021 , 28 ( 1 ): 66 - 75 .
Chen C , Wang N , Chen M . Prediction model of end-point phosphorus content in consteel electric furnace based on PCA-extra tree model [J]. ISIJ International , 2021 , 61 ( 6 ): 1908 - 1914 .
He F , Zhang L Y . Prediction model of end-point phosphorus content in BOF steelmaking process based on PCA and BP neural network [J]. Journal of Process Control , 2018 , 66 : 51 - 58 .
Zhou K X , Lin W H , Sun J K , et al . Prediction model of end-point phosphorus content for BOF based on monotone-constrained BP neural network [J]. Journal of Iron and Steel Research International , 2022 , 29 ( 5 ): 751 - 760 .
Liu Z , Cheng S S , Liu P B . Prediction model of BOF end-point P and O contents based on PCA-GA-BP neural network [J]. High Temperature Materials and Processes , 2022 , 41 ( 1 ): 505 - 513 .
Zou Y C , Yang L Z , Li B , et al . Prediction model of end-point phosphorus content in EAF steelmaking based on BP neural network with periodical data optimization [J]. Metals , 2022 , 12 ( 9 ): 1519 .
Bashir T , Chen H Y , Tahir M F , et al . Short term electricity load forecasting using hybrid prophet-LSTM model optimized by BPNN [J]. Energy Reports , 2022 , 8 : 1678 - 1686 .
Samantaray S , Sahoo A . Prediction of runoff using BPNN, FFBPNN, CFBPNN algorithm in arid watershed: A case study [J]. International Journal of Knowledge-Based and Intelligent Engineering Systems , 24 ( 3 ): 243 - 251 .
Li H , Barui S , Mukherjee S , et al . Least squares twin support vector machines to classify end-point phosphorus content in BOF steelmaking [J]. Metals , 2022 , 12 ( 2 ): 268 .
Chandra M A , Bedi S S . Survey on SVM and their application in imageclassification [J]. International Journal of Information Technology , 2021 , 13 ( 5 ): 1 - 11 .
Kurani A , Doshi P , Vakharia A , et al . A comprehensive comparative study of artificial neural network (ANN) and support vector machines (SVM) on stock forecasting [J]. Annals of Data Science , 2023 , 10 ( 1 ): 183 - 208 .
Liang Y R , Wang H B , Xu A J , et al . A two-step case-based reasoning method based on attributes reduction for predicting the endpoint phosphorus content [J]. ISIJ International , 2015 , 55 ( 5 ): 1035 - 1043 .
Xin Z C , Zhang J S , Jin Y , et al . Predicting the alloying element yield in a ladle furnace using principal component analysis and deep neural network [J]. International Journal of Minerals, Metallurgy and Materials , 2023 , 30 ( 2 ): 335 - 344 .
Aldahdooh A , Hamidouche W , Fezza S A , et al . Adversarial example detection for DNN models: A review and experimental comparison [J]. Artificial Intelligence Review , 2022 , 55 ( 6 ): 4403 - 4462 .
Devan P , Khare N . An efficient XGBoost-DNN-based classification model for network intrusion detection system [J]. Neural Computing and Applications , 2020 , 32 ( 16 ): 12499 - 12514 .
Sainburg T , McInnes L , Gentner T Q . Parametric UMAP embeddings for representation and semisupervised learning [J]. Neural Computation , 2021 , 33 ( 11 ): 2881 - 2907 .
Tikhamarine Y , Souag-Gamane D , Najah Ahmed A , et al . Improving artificial intelligence models accuracy for monthly streamflow forecasting using grey Wolf optimization (GWO) algorithm [J]. Journal of Hydrology , 2020 , 582 : 124435 .
邹雷雷 , 吕 庆 , 刘东辉 , 等 . 基于加权灰色关联度法分析印尼海砂最佳配比的研究 [J]. 中国冶金 , 2014 , 24 ( 9 ): 6 - 9 .
Zhang S S , Gao H , Wei G W , et al . Grey relational analysis method based on cumulative prospect theory for intuitionistic fuzzy multi-attribute group decision making [J]. Journal of Intelligent & Fuzzy Systems , 2021 , 41 ( 2 ): 3783 - 3795 .
Lv H L , Chen X Y , Wang X Y , et al . A novel study on a micromixer with Cantor fractal obstacle through grey relational analysis [J]. International Journal of Heat and Mass Transfer , 2022 , 183 : 122159 .
Kobak D , Linderman G C . Initialization is critical for preserving global data structure in both t-SNE and UMAP [J]. Nature Biotechnology , 2021 , 39 ( 2 ): 156 - 157 .
Wang Y F , Huang H Y , Rudin C , et al . Understanding how dimension reduction tools work: An empirical approach to deciphering t-SNE, UMAP, TriMAP, and PaCMAP for data visualization [J]. J Mach Learn Res , 2016 , 22: 201: 1 - 201 : 73 .
Dubey A , Mishra R , Cheng C W , et al . Demonstration of a superior deep-UV surface-enhanced resonance Raman scattering (SERRS) substrate and single-base mutation detection in oligonucleotides [J]. Journal of the American Chemical Society , 2021 , 143 ( 46 ): 19282 - 19286 .
Tembhurne J V , Almin M M , Diwan T . Mc-dnn [J]. International Journal on Semantic Web and Information Systems , 2022 , 18 ( 1 ): 1 - 20 .
李勃 , 杨凌志 , 等 . 9电弧炉炼钢流程一键合金加料优化系统应用 [J]. 钢铁 . 2022 , 57 ( 4 ): 58 - 67 .
0
浏览量
0
下载量
0
CSCD
0
CNKI被引量
关联资源
相关文章
相关作者
相关机构
京公网安备11010802024621