Yujie Zhang

Yujie Zhang

Auto-generated profile, based on publicly available information

Yujie Zhang

North China University of Science and Technology

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
18 February 2026 in Journal of Sustainable Metallurgy
Verified
Journal Article
Closed access
PDF file
YZYujie Zhang
Yujie Zhang
XLXiaojie Liu
Xiaojie Liu
RLRan Liu
Ran Liu
HLHongwei Li
Hongwei Li
YDYifan Duan
Yifan Duan
LLLanjie Li
Lanjie Li

8 October 2026 in Ironmaking & Steelmaking: Processes, Products and Applications
Verified
Journal Article
Closed access
WBWang Bo
Wang Bo
XLXiaojie Liu
Xiaojie Liu
RLRan Liu
Ran Liu
ZYZhang Yujie
Zhang Yujie
LXLi Xin
Li Xin
HLHongwei LI
Hongwei LI
The blast furnace is a high-temperature closed black-box reaction vessel. Its internal physical and chemical processes are characterized by strong coupling, large time delay, and multivariable heterogeneous time lags, which severely restrict the accurate prediction and closed-loop control of hot metal output. Based on real industrial data, this paper proposes a complete time-lag feature-based attention-Mamba collaborative scheme for blast furnace hot metal output prediction and control. First, an adaptive correction self-attention-based imputation for time series adaptive anomaly correction model is constructed, which accurately quantifies the dominant time lags and effective action intervals of each process parameter on hot metal output by virtue of the diagonal masked causal attention mechanism. Then, a comprehensive importance evaluation index is extracted based on the attention weight distribution to screen core feature parameters. On this basis, an attention-Mamba hot metal output prediction model is proposed, which integrates the global attention importance vector as prior knowledge into the selective scanning mechanism of Mamba. Finally, an attention-Mamba-based genetic algorithm-model predictive control multivariable collaborative control framework is established, which deeply embeds the extracted time-lag parameters into the control logic and achieves global optimization under physical safety constraints and state violation penalty mechanisms. The experimental results show that: (1) the dataset processed with adaptive correction self-attention-based imputation for time series time-delay weighted alignment significantly reduces the root mean square error (RMSE) of the proposed attention-Mamba model by 37.2% compared to the original unaligned data (from 14.215 to 8.923 t); (2). The single-step prediction RMSE of the attention-Mamba model is as low as 8.923 t, with an R 2 of 0.931 and a ±15 t prediction hit rate of 93.2%; (3) industrial field tests demonstrate that the proposed control framework reduces the average error of hot metal output per cast by 38.9% and the maximum error by 48.2%. This study realizes the deep integration of high-precision hot metal output prediction and multivariable collaborative control for blast furnaces, and provides reliable theoretical and technical support for the intelligent closed-loop control of the blast furnace ironmaking process.

7 September 2026 in Ironmaking & Steelmaking: Processes, Products and Applications
Verified
Journal Article
Closed access
XLXiaojie Liu
Xiaojie Liu
XSXinxin Su
Xinxin Su
HLHongwei LI
Hongwei LI
HLHong-yang Li
Hong-yang Li
YZYujie Zhang
Yujie Zhang
XLXin Li
Xin Li
To address the key challenges in blast furnace ironmaking, such as the strong coupling among multiple molten iron quality variables and significant nonlinearity, this paper constructs an intelligent whole-process control method that integrates prediction, evaluation, and optimisation decision-making. Based on actual blast furnace production data, a feature selection framework that accounts for both nonlinear correlations and causal mechanisms is established by combining Hilbert-Schmidt Independence Criterion (HSIC)-Lasso and the Peter-Clark Momentary Conditional Independence (PCMCI) algorithm, effectively eliminating spurious feature associations. Using the selected features, a TimeMixer multi-scale time-series prediction model is constructed to achieve high-precision prediction of four core quality indicators: molten iron temperature (MIT), molten iron silicon content ([Si]), molten iron phosphorus content ([P]), and molten iron sulphur content ([S]). The root mean square errors are reduced by more than 12.5% on average compared with traditional time-series models. On this basis, the prediction results are used to build an improved TOPSIS evaluation model that integrates game theory with subjective and objective combined weighting, and K-Means++ clustering is applied to achieve quantitative characterisation and four-level grading of molten iron quality. Finally, using the evaluation results as decision criteria, a Bayesian optimisation control model is further constructed to efficiently search for optimal control parameters under the dual constraints of process hard limits and operational change rates. Industrial application and validation results show that the qualification rates of the core targets MIT and [Si] are significantly improved from 71% and 75% to 85% and 89%, respectively; the comprehensive evaluation score of molten iron quality increases from 0.768 to 0.864; and the quality grade is upgraded from ‘Good’ to ‘Excellent’. This method not only greatly enhances the ability to meet key quality indicators but also provides an innovative integrated solution for intelligent control of the blast furnace ironmaking process.

1 January 2026 in Journal of Mining and Metallurgy, Section B: Metallurgy
Verified
Journal Article
Open access
PDF file
XLXiaojie Liu
Xiaojie Liu
YGYanping Geng
Yanping Geng
HLHongwei LI
Hongwei LI
YZYujie Zhang
Yujie Zhang
YDYifan Duan
Yifan Duan
XLXin Li
Xin Li
Accurate prediction of blast furnace conditions and traceability of anomaly root causes are crucial for ensuring production stability and reducing energy consumption. However, existing methods struggle to simultaneously capture temporal dynamic evolution and causal structural relationships among parameters. To address this issue, this paper proposes a method for furnace condition prediction and root cause tracing that integrates a knowledge graph with dual-encoder deep learning. First, based on process mechanisms and expert rules, a knowledge graph containing anomaly types and their causal relationships is constructed, and historical cases are stored. Second, a temporal-graph dual encoder is designed: the temporal encoder employs a temporal convolutional network to extract dynamic evolution features from data, while the graph encoder uses a graph attention network to learn structural dependencies from the knowledge graph; a gated fusion mechanism then enables accurate prediction of six types of furnace conditions. Finally, a root cause tracing algorithm based on causal paths and case similarity is proposed, which leverages causal associations in the knowledge graph and historical cases to generate interpretable root cause chains. Experimental results show that the proposed method achieves an overall prediction accuracy of over 90% and an overall tracing accuracy of 91.7%, and that the generated paths are highly consistent with process mechanisms.

28 April 2026 in Journal of Sustainable Metallurgy
Verified
Journal Article
Closed access
PDF file
ZMZhang Mengjie
Zhang Mengjie
RLRan Liu
Ran Liu
XLXiaojie Liu
Xiaojie Liu
ZYZhang Yujie
Zhang Yujie
LXLi Xin
Li Xin
HLHongwei LI
Hongwei LI

8 October 2026 in Ironmaking & Steelmaking: Processes, Products and Applications
Verified
Journal Article
Closed access
RLRan Liu
Ran Liu
BWBo Wang
Bo Wang
XLXiaojie Liu
Xiaojie Liu
YDYifan Duan
Yifan Duan
YZYujie Zhang
Yujie Zhang
HLHongwei LI
Hongwei LI
To address the challenges of strong temporal dependencies, multivariable coupling, and nonlinearity in blast furnace hot metal output prediction, this study proposes an interpretable prediction framework that integrates a feature selection module based on Light Gradient Boosting Machine and Recursive Feature Elimination with Cross-Validation (LightGBM_RFECV) with a hybrid LSTM-Transformer model. First, an iterative two-stage feature optimisation framework is adopted: feature importance is evaluated using LightGBM (based on information gain), and redundant features are eliminated through RFECV with recursive cross-validation. Then, a combined prediction model that integrates temporal modelling and attention mechanisms (LSTM-Transformer) is constructed, achieving high prediction accuracy (MAE = 5.47, R 2 = 0.8747, and a ± 20-ton accuracy rate of 96.17%), significantly outperforming baseline models. Furthermore, the Captum toolkit is introduced to perform feature attribution, quantifying the influence mechanisms of process variables on output and forming an intelligent ‘prediction–attribution–regulation’ closed loop. Experiments based on real production data verify the effectiveness of the proposed method, providing data-driven decision support for energy conservation, emission reduction, and stable blast furnace operation, thereby contributing to the intelligent transformation of the steel industry.