Association Rules Mining for Oxygen Consumption Reduction in Basic Oxygen Furnace Steelmaking Process

The basic oxygen furnace (BOF) steelmaking process in the steel industry is highly complicated, and subject to variations in raw material composition. During the BOF steelmaking process, it is essential to maintain the carbon content and the endpoint temperature at their set points in the liquid steel. This paper presents intelligent models used to estimate the endpoint temperature in the basic oxygen furnace (BOF) steelmaking process. An artificial neural network (ANN) model and a least-squares support vector machine (LSSVM) model are proposed and their estimation performance compared. The classical partial least-squares (PLS) method was also compared with the others. Results of the estimations using the ANN, LSSVM and PLS models were compared with the operation data, and the root-mean square error (RMSE) for each model was calculated to evaluate estimation performance. The RMSE of the LSSVM model 15.91, which turned out to be the best estimation. RMSE values for the ANN and PLS models were 17.24 and 21.31, respectively, indicating their relative estimation performance. The essential input parameters used in the models can be selected by sensitivity analysis. The RMSE for each model was calculated again after a sequential input selection process was used to remove insignificant input parameters. The RMSE of the LSSVM was then 13.21, which is better than the previous RMSE with all 16 parameters. The results show that LSSVM model using 13 input parameters can be utilized to calculate the required values for oxygen volume and coolant needed to optimally adjust the steel target temperature. (Received July 30, 2018; Accepted September 28, 2018)

Black box operation optimization of basic oxygen furnace steelmaking process with derivative free optimization algorithm

Basic Oxygen Furnace (BOF) steelmaking is the main method for steelmaking production, mixing efficiency of BOF vessel would greatly affect the uniformity of chemical composition and temperature in molten steel. Therefore, this paper conducted the effect of blowing parameters on bath mixing efficiency during BOF steelmaking process. Through measuring the mixing time of steel bath in BOF vessel, the effect of different blowing parameters was studied. Based on analysis of variance for the experimental results, the influence degree on mixing efficiency by different blowing parameters was acquired, and the optimized combination of blowing parameters was obtained. Moreover, the mixing index which indicated the mixing degree of steel bath in BOF vessel was proposed in this study, and an apparent nonlinear relation was acquired after determining relationship between the mixing index and blowing parameters. The optimal blowing parameters were applied to an 80 t industrial BOF vessel in a special steel plant, the end point carbon and oxygen equilibrium was decreased from 0.0032 to 0.0026, end point phosphorus and total Fe content in slag were reduced by 0.0012% and 2.21%, respectively. The mixing index and results presented in this work could be helpful to improve mixing efficiency in commercial BOF vessels.

This article studies an operation optimization problem in a steelmaking process. Shortly before the tapping of molten steel from the basic oxygen furnace (BOF), end-point control measures are applied to achieve the required final molten steel quality. While it is difficult to build an exact mathematical model for this process, the control inputs and the corresponding outputs are available by collecting production data. We build a data-driven model for the process. To optimize the control parameters, an improved estimation of distribution algorithm (EDA) is developed using a probabilistic model comprising different distributions. A resampling mechanism is incorporated into the EDA to guide the new population to a broader and more promising area when the search becomes ineffective. To further enhance the solution quality, we add a local improvement to update the current best individual through simplified gravitational search and information learning. Experiments are conducted using real data from a BOF steelmaking process. The results show that the algorithm can help to achieve the specified molten steel quality. To evaluate the proposed algorithm as a general optimization algorithm, we test it on some complex benchmark functions. The results illustrate that it outperforms other state-of-the-art algorithms across a wide range of problems.

A 3D mathematical model on the multiphase flow and splashing phenomenon in a 210 t Basic oxygen furnace converter is established. The turbulent multiphase flow of the molten steel, slag, and oxygen is simulated by the combined realizable k−ε model and volume of fluid model. The trajectory of argon bubbles and the interaction between the molten steel and argon is simulated by the two‐way coupled Lagrangian discrete phase model. Effects of the lance height and bottom‐blowing gas flow rate on the multiphase flow and splashing phenomenon are investigated. The results indicated that the cavity depth is decreased from 0.45 to 0.36 m when the lance height increases from 1700 to 2300 mm. The cavity depth is not significantly affected by the bottom‐blowing argon flow rate with the 44 000 m3 h−1 top‐blowing oxygen flow rate and 1700 mm lance height. The increased lance height resulted in a decrease in splashing. When the oxygen lance height increases from 1700 to 2300 mm, the amount of the splashed molten steel decreases from 3774.3 to 1953.0 kg.

A significant amount of CO2 is emitted and utilized in the production processes engaged by iron and steel enterprises. Herein, the feasibility and principles of CO2 participation in the steelmaking process are analyzed, and an industrial examination of bottom‐blown CO2 smelting in a 120 t converter is performed; this serves to summarize and verify the metallurgical advantages of CO2 participation in the converter steelmaking process. The injection of CO2 improves the effects of dephosphorization and denitrification, reduces the end‐point oxygen content of molten steel, diminishes the consumption of the slagging agent, and contributes to a reduction in production costs and improvement in product quality. This study analyzes two different recycling paths of CO2 in iron and steel enterprises, providing a new mechanism for CO2 emission reduction.

This study investigates the application of advanced deep learning models, particularly transformer-based architectures, to model the basic oxygen furnace (BOF) steelmaking process, with a focus on accurately predicting endpoint temperature. Achieving precise temperature predictions is critical for optimizing process parameters, enhancing product quality, and ensuring energy efficiency in BOF operations. This research represents the first extensive deployment of such models on BOF operational data. We rigorously evaluated five state-of-the-art deep learning models using a comprehensive dataset consisting of over 10,000 samples from a large-scale industrial setting. Among these models, we introduced a novel architecture, TTB-BOFNet, tailored to handle the complexities of industrial-scale BOF data. TTB-BOFNet leverages transformer models enhanced through exploratory data analysis-informed binning to better interpret intricate interactions and generate contextual embeddings. Our novel architecture demonstrated a notable improvement in predictive performance, achieving an 8.81 pct improvement in R-squared over XGBoost, a 4.65 pct gain over the multi-layer perceptron (MLP) benchmark and a 1.79 pct improvement over the TabTransformer model. Additionally, SHapley Additive exPlanations (SHAP) analysis was employed to interpret the TTB-BOFNet model predictions, revealing the effectiveness of both the EDA-informed binning approach and the model’s transformer-based structure in capturing critical features and improving predictive accuracy. These findings underscore the potential of transformer-based deep learning models to optimize complex industrial operations. TTB-BOFNet offers a robust, interpretable, and scalable solution for BOF temperature prediction, paving the way for more adaptive, data-driven decision-making in steel manufacturing.

Different mixed cultures of extremely thermoacidophilic microorganisms were used for bioleaching of metalliferous industrial dust waste derived from the basic oxygen furnace (BOF) steelmaking process. Such mixed cultures can extract various metals from multi-metallic BOF-dust waste, improving the metal dissolution and bioleaching performance in frames of metal recycling processes to assist circular economies and waste management. The results of the investigation showed that mixed cultures of thermoacidophilic archaea of the order Sulfolobales (Acidianus spp., Sulfolobus spp., and Metallosphaera sedula) during their growth in laboratory glass bioreactors provided a superior bioleaching system to Acidianus manzaensis alone. Depending on the composition of mixed thermoacidophilic cultures, extraction of various metals from BOF-dust could be achieved. Among the three different types of mixed cultures tested, the mixed culture system of A. manzaensis, A. brierleyi, and S. acidocaldarius was most effective for extraction of major elements (Fe, Ca, Zn, Mn, and Al). The mixed culture of A. manzaensis, A. brierleyi, and M. sedula showed high performance for bioleaching of most of the minor elements (Cu, Ni, Pb, Co, Mo, and Sr). The efficient ability of mixed cultures to colonise the mineral matrix of the metal waste product was observed via scanning electron microscopy, while their metal extraction capacities were analysed by inductively coupled plasma mass spectrometry. These investigations will promote the further design of microbial consortia in order to break down the solid matrix and efficiently extract metals from metalliferous waste materials.

In this study, the decarburization behavior in basic oxygen furnace (BOF) steelmaking was investigated based on the multi-zone reaction mechanism. The contributions of the main reaction zones to decarburization were clarified, and the effects of key factors—including the effective reaction amount in the main reaction zones, the post combustion ratio (PCR) in auxiliary reaction zones, and the carbon content of scrap steel—on decarburization behavior were quantitatively analyzed. The results indicate that decarburization predominantly occurs in the jet impact reaction zone (approximately 76% of the total decarburization), followed by the emulsion and metal droplet reaction zone (approximately 14%) and the bulk metal and slag reaction zone (approximately 10%). Variations in the effective reaction amount for the main reaction zones significantly affect both the decarburization rate and the endpoint carbon content, with the direct oxidation decarburization reaction in the jet impact reaction zone being the dominant factor. In addition, the PCR in the gas homogenization zone of the auxiliary reaction zones determines the distribution ratio of effective reaction oxygen, while the melting behavior of scrap steel in the metal homogenization zone plays a critical role in the precise control of the endpoint carbon content. This study provides a quantitative elucidation of the effects of different reaction zones on decarburization behavior, offering a foundation for the precise control of endpoint carbon content in BOF steelmaking.

The endpoint carbon content of steelmaking is an important criterion for steel quality. Aiming at increasing the accuracy of endpoint carbon content prediction in basic oxygen furnace (BOF) steelmaking, this paper uses case-based reasoning (CBR) method to predict the endpoint carbon content of BOF steelmaking. In CBR, case retrieval makes a significant impact on reasoning result. Therefore, we apply affinity propagation (AP) clustering algorithm and waterfilling algorithm to enhance the case retrieval so as to improve the accuracy and stability of endpoint carbon content prediction. Through the simulation experiment, this paper compares the new model we proposed with the widely used method at present. The results show that the improved CBR can obviously improve the accuracy of endpoint carbon content prediction.

Traditional basic oxygen furnace (BOF) prediction based on Case-based reasoning (CBR) is usually dependent on experts and experiences which will result in deviation. Aiming at attributes selection in BOF steelmaking and case base maintenance, this paper proposes using Ridge regression and Association rules to improve CBR and predicting the endpoint carbon content of BOF steelmaking. In CBR, case representation is the basic, the selected attributes in case representation play an important role but dependent on expert experiences so that this paper uses lasso regression algorithm to get the attributes. Through the simulation experiment, the results show that the improved CBR can obviously improve the accuracy of endpoint carbon content prediction.

ABSTRACTThe influence of the addition amount of conditioning agent, gas quenching pressure, nozzle type and slag temperature on gas quenching effect of basic oxygen furnace (BOF) steelmaking slag were studied through the gas quenching system. The results show that the gas quenching rate basically increased with the increase in the amount of the conditioning agent by using BF slag as a conditioning agent. High gas quenching pressure and high Mach number of nozzle had a slight promotion effect on gas quenching rate. While the modified slag temperature had a significant influence on the gas quenching rate. The higher the slag temperature and the superheat degree, the higher the gas quenching rate. Gas-quenched steel slag micro-beads property is related to the amount of the conditioning agent. The granulation effect became worse and the mineral phases gradually changed to the glass phase with the increasing amount of the conditioning agent.

The endpoint control of Basic Oxygen Furnace (BOF) steelmaking depends on the prediction of the endpoint carbon content and temperature. However, predicting these variables is challenging because of the numerous working conditions in the industrial field and the volatility of the sensor data collected during BOF steelmaking. The accuracy of prediction models in ensemble learning depends significantly on the initial distribution of data. However, the complex nature of BOF steelmaking data makes it challenging to generate diverse subsets, which ultimately affects the accuracy of predictions. This paper presents a new approach called Graph Convolutional Network Node Embedding Supervised Ensemble Clustering (GESupEC) for soft sensor modelling in ensemble learning to tackle these issues. GESupEC utilizes a similarity graph derived from a co-association matrix and employs graph convolutional networks to extract structural information among nodes. By optimising the clustering loss within the network, GESupEC learns compact node representations that are useful for the clustering task. Furthermore, it generates a reconstruction matrix based on the similarity of node embeddings. This matrix helps with the extraction of a suitable subset of data for BOF steelmaking through matrix decomposition. After that, the gradient boosting decision tree regression sub-model is established based on the data subset. An ensemble strategy called Gray Relational Analysis Weighted Average is proposed, which assigns weights based on the grey relation similarity between test samples and different data subsets. This weighted average strategy aims to enhance the accuracy of carbon content and temperature predictions. When tested with actual BOF steelmaking generation process data, the prediction accuracy of carbon content reached 88.6% within the error range of ±0.02%, and the prediction accuracy of temperature reached 92.6% within the error range of ±10 °C.

The research focuses on the accurate prediction of end-point carbon content and temperature in basic oxygen furnace (BOF) steelmaking. This article proposes a static prediction method based on PSO-optimised KNN-weighted twin support vector regression (KPCA-KNNWTSVR). Firstly, key input variables, such as initial hot metal conditions, raw materials, and operational parameters, are selected through grey relational analysis to eliminate redundancy. Subsequently, Kernel Principal Component Analysis (KPCA) is utilised to extract nonlinear features and suppress noise. Simultaneously, sample weights are calculated based on the K-nearest neighbours (KNN) algorithm to enhance local fitting capability. The feature vectors and weights are then integrated into the weighted twin support vector regression (KNNWTSVR) model. Furthermore, the PSO algorithm is employed to adaptively search for the model's hyperparameters to further improve its generalisation performance. Simulation and experimental results indicate that the proposed method outperforms traditional models in both prediction accuracy and stability. With allowable errors of ±0.01 wt-% for carbon content and ±15°C for temperature. The model achieved hit rates of 80% and 88% on 100 actual production datasets. Compared with the Support Vector Machine (SVM) model, the hit rate for carbon content prediction is improved by 5 percentage points. Compared with the backpropagation (BP) neural network model, the hit rate for temperature prediction is improved by 20 percentage points. This method performs best among all compared approaches.

A static end-point prediction model based on automatic optimization is proposed to predict the end-point carbon content and temperature of the basic oxygen furnace (BOF) steelmaking. Initially, redundant features in the samples are eliminated through mechanistic analysis, and violin plots are employed to minimize the impact of outliers. Subsequently, Spearman correlation analysis is utilized to identify statistically significant features influencing the end-point, hence constructing a highly interpretable feature set. Finally, leveraging the advantages of the lévy-flying algorithm (LWOA), which offers simplicity in parameter tuning and strong global search capabilities, the parameters of the projection wavelet weighted twin support vector regression (PWWTSVR) objective function are optimized. These parameters include the regularization parameter c 1 , c 2 , c 3 , c 4 , c 5 , c 6 , the insensitivity parameter ε 1 , ε 2 , the width of the radial basis function σ , the amplitude E , and the standard deviation γ . Additionally, by setting different population sizes (1, 20, 50, 100), the influence of the whale population size in LWOA on the predictive model is investigated. Simulation results show that the prediction model achieves hit rates of 94% for end-point carbon content and 97% for temperature, with error tolerances of ±0.005 wt.% and ±10 °C, respectively. The dual hit rate stands at 91%, outperforming five other models. Furthermore, the predictive capability of the proposed model is comprehensively evaluated by implementing various error intervals, hence validating its effectiveness.