Molten Iron Quality Research Articles

Research on Molten Iron Quality Prediction Based on Machine Learning

The quality of molten iron not only has a significant impact on the strength, toughness, smelting cost and service life of cast iron but also directly affects the satisfaction of users. The establishment of timely and accurate blast furnace molten iron quality prediction models is of great significance for the improvement of the production efficiency of blast furnace. In this paper, Si, S and P content in molten iron is taken as the important index to measure the quality of molten iron, and the 989 sets of production data from a No.1 blast furnace from August to October 2020 are selected as the experimental data source, predicting the quality of molten iron by the I-GWO-CNN-BiLSTM model. First of all, on the basis of the traditional data processing method, the missing data values are classified into correlation data, temporal data, periodic data and manual input data, and random forest, the Lagrangian interpolation method, the KNN algorithm and the SVD algorithm are used to complete them, so as to obtain a more practical data set. Secondly, CNN and BiLSTM models are integrated and I-GWO optimized hyperparameters are used to form the I-GWO-CNN-BiLSTM model, which is used to predict Si, S and P content in molten iron. Then, it is concluded that using the I-GWO-CNN-BiLSTM model to predict the molten iron quality can obtain high prediction accuracy, which can provide data support for the regulation of blast furnace parameters. Finally, the MCMC algorithm is used to analyze the influence of the input variables on the Si, S and P content in molten iron, which helps the steel staff control the quality of molten iron in a timely manner, which is conducive to the smooth running of blast furnace production.

Long Short-Term Memory Parameter Optimization Based on Improved Sparrow Search Algorithm for Molten Iron Quality Prediction

Blast furnace (BF) ironmaking is a key process in iron and steel production. Because BF ironmaking is a dynamic time series process, it is more appropriate to use a recurrent neural network for modeling. The long short-term memory (LSTM) network is commonly used to model time series data. However, its model performance and generalization ability heavily depend on the parameter configuration. Therefore, it is necessary to study parameter optimization for the LSTM model. The sparrow search algorithm (SSA) holds advantages over traditional optimization algorithms in several aspects, such as no need for prior knowledge, fewer parameters, fast convergence, and high scalability. However, the algorithm still faces some challenges, such as the tendency to become trapped in the local optimum and the imbalance between global search ability and local search ability. Therefore, on the basis of SSA, this study examined the Levy flight strategy, sine search strategy, and step size factor adjustment strategy to improve it. This algorithm, improved by three strategies, is called the improved sparrow search algorithm (ISSA). Then, the ISSA-LSTM model was established. Furthermore, considering the limitations of SSA in dealing with multi-objective problems, the fast non-dominated sorting genetic algorithm (NSGAII) was introduced, and the ISSA-NSGAII model was established. Finally, experimental validation was performed using real blast furnace operation data, which demonstrated the proposed algorithm’s superiority in parameter optimization for the LSTM model and prediction for real industrial data.

Optimal Tracking Control of Blast Furnace Molten Iron Quality Based on Krotov's Method and Nonlinear Subspace Identification

Optimal Tracking Control of Blast Furnace Molten Iron Quality Based on Krotov's Method and Nonlinear Subspace Identification

One-Sided Relational Autoencoder With Seasonal-Trend Decomposition to Extract Process Correlations for Molten Iron Quality Prediction

One-Sided Relational Autoencoder With Seasonal-Trend Decomposition to Extract Process Correlations for Molten Iron Quality Prediction

A Cooperative Silicon Content Dynamic Prediction Method With Variable Time Delay Estimation in the Blast Furnace Ironmaking Process

Online assessment of the molten iron quality of a blast furnace ironmaking process strongly depends on reliable measurement of silicon content (Si). This article proposes a novel cooperative strategy to train the data-driven prediction model and estimate variable time delay (VTD) values. First, a dynamic deep network that stacks denoising autoencoders (D-SDAE) with the ability to describe process nonlinearity and time-varying is designed for Si estimation. Then, time delay between the process variables and Si, which may disrupt the data distribution pattern (i.e., the real input-output relationship), is recovered in the modeling process. It is an overlooked data feature that occurs due to material transportation times, installation location and the analysis period of sampling equipment. The VTD values are considered to be the parameters of the D-SDAE model and obtained in the training process by the proposed double-scale collaborative search particle swarm optimization algorithm. The effectiveness of the proposed VTD-based D-SDAE model is validated in a numerical simulation and an industrial ironmaking plant, and a marked improvement in the prediction performance is achieved when VTD information is considered.

Optimal tracking control for blast furnace molten iron quality based on subspace identification and Krotov's method

SummaryBlast furnace ironmaking is a complex nonlinear dynamic process, which has strict requirements on the quality of final molten iron. To improve the control effect of molten iron quality, this paper proposes an optimal tracking control method for molten iron quality based on the bilinear subspace identification and Krotov's method. First, a state space model between key process variables and blast furnace molten iron quality, using the bilinear subspace identification method, is established based on actual industrial data. Then, an augmented system is established combining the state vector and the reference trajectory, and the improving feedback controller is designed by solving a set of Riccati equations, which are established based on Krotov's method. In combination with the designed controller and the established system, the specific implementation steps are proposed. Finally, the effectiveness and advancement of the proposed method are verified based on actual industrial data of a large blast furnace.

Prediction of Multiple Molten Iron Quality Indices in the Blast Furnace Ironmaking Process Based on Attention-Wise Deep Transfer Network

Molten iron quality (MIQ) indices prediction based on data-driven models is an important way to monitor product quality and smelting status in the blast furnace ironmaking process. However, some challenges still place in the MIQ prediction: 1) limited nonlinear and dynamic description capabilities and interpretability of data-driven models; 2) high demand on the number of the labeled samples; 3) insufficient exploration of the underlying relationship between MIQ indices. In this case, we propose a novel data-driven deep model for the online prediction of MIQ indices. First, we design an attention-wise module to self-learn the nonlinear and dynamic relationship between process variables and prediction targets and enhance interpretability. Then, the minute-level molten iron temperature data detected by our previously developed equipment is used to pre-train the attention-wise deep network to obtain the improved weights and reduce dependence on labeled samples. Finally, the pre-trained model is extended to a structure with a weight-shared attention-wise module and task-separated prediction networks to explore the relationship between multiple prediction tasks. The effectiveness of the proposed attention-wise deep network is verified in an industrial ironmaking plant, which shows a significant improvement in performance, i.e., high accuracy and interpretability.

Temporal Hypergraph Attention Network for Silicon Content Prediction in Blast Furnace

Online dynamic prediction of the hot metal silicon content in the blast furnace ironmaking process is crucial for stabilizing the furnace condition and improving the molten iron quality. However, due to the complex nonlinear correlations and time-varying time lags between silicon content and process variables, silicon content prediction is a challenging task. To tackle the problem, we propose a novel silicon content prediction method, called temporal hypergraph attention network (T-HyperGAT), which is combined the hypergraph attention network (HyperGAT) and the gated recurrent unit (GRU) network. Specifically, the Hyper-GAT is used to capture the high-order correlations of input features and perform equal-dimensional feature transformation to maintain the temporality of input features, and the GRU network is used to capture the time-series characteristics of transformed input features to overcome the time-varying time lags of process variables. Then the T-HyperGAT model can capture high-order correlations and time-series characteristics from complex industrial data. The effectiveness of the proposed T-HyperGAT method is verified by actual blast furnace ironmaking process data from a blast furnace in China.

Quality-related process monitoring of ironmaking blast furnace based on improved kernel orthogonal projection to latent structures

This paper proposes a novel improved kernel orthogonal projection to latent structures (KOPLS) for quality-related process monitoring of ironmaking blast furnace (BF). Firstly, the KOPLS algorithm is used as a preprocessing tool to remove the components which are not related to the molten iron quality (MIQ) in the process variable space. Then, the remaining process variable space is further orthogonally decomposed into the quality-related subspace and the quality-unrelated subspace by the post-processing method. Afterwards, the complete decomposition of the process variable space is achieved and each subspace has obvious correlation or irrelevance with MIQ. By designing appropriate monitoring statistical indicators in each subspace, it can clearly detect whether the fault is related to the MIQ. Aiming at the problem of fault identification based on the improved KOPLS, this paper further proposes a fault identification method based on robust reconstruction error. This method reversely estimates the normal values of the process variables in the original variable space based on the improved KOPLS model, and uses the reconstruction error of the process variables to construct the fault identification index. Numerical experiments and industrial experiments all verify the effectiveness and practicability of the proposed method.

Output space transfer based multi-input multi-output Takagi–Sugeno fuzzy modeling for estimation of molten iron quality in blast furnace

The molten iron quality (MIQ) in the blast furnace (BF) needs on-site sampling and laboratory tests, which is very unfavorable to the control. To estimate multiple molten iron quality indicators (MIQIs) such as molten iron temperature (MIT), silicon content ([Si]), phosphorus content ([P]) and sulfur content ([S]) in real-time, some data-driven multi-input multi-output (MIMO) modeling methods have lately been developed. However, those modeling methods ignore the inter-indicator correlation. Moreover, the above modeling methods do not consider the processing strategy for the samples with missing indicators. To conquer the above two problems in the process of modeling, this paper proposed an output space transfer based MIMO Takagi–Sugeno (T–S) fuzzy modeling method for the four MIQIs. Through sparse learning of the correlation matrix, the inter-indicator correlation is specifically modeled. Due to the full use of inter-indicator correlation, the impact of missing indicators on modeling accuracy is greatly reduced. An effective alternative optimization algorithm is also provided for the corresponding intractable optimization problem. The actual data test shows that the proposed method can greatly improve the accuracy of multiple MIQIs estimations.

Fast just-in-time-learning recursive multi-output LSSVR for quality prediction and control of multivariable dynamic systems

Aiming at quality prediction and control of blast furnace (BF) ironmaking process characterized by complicated nonlinear time-varying dynamics, this paper proposes a just-in-time-learning (JITL) recursive multi-output least squares support vector regression (JITL-R-M-LSSVR) algorithm with fast nonlinear local learning capability for multivariable dynamic systems. The proposed fast JITL-R-M-LSSVR effectively combines the online local learning of JITL with the multi-output LSSVR (M-LSSVR) based on multi-task transfer learning, and focuses on how to ensure the rapid verification of the local model during online learning of M-LSSVR, and how to perform model pruning while recursively updating the model parameters to improve the calculation efficiency. To this end, the proposed algorithm uses a derived multi-output incremental learning algorithm to recursively update model parameters online in a gentle way, which has better modeling stability and smoothness than the traditional way that discards old models. At the same time, when the model is pruned, a novel multi-output reverse decremental learning algorithm is proposed to adaptively delete the modeling data, so as to effectively control the sample size and reduces the calculation cost. In particular, the model verification of the proposed algorithm only needs to construct the M-LSSVR modeling matrix and the matrix inverse operation once, and the matrix after deleting each modeling sample can be easily obtained by reverse decremental learning of the original modeling matrix, which can achieve fast and efficient model verification. Finally, the effectiveness and practicability of the proposed method are verified by applying it to prediction modeling and predictive control of the molten iron quality in BF ironmaking process.

Kalman Filter-Based Data-Driven Robust Model-Free Adaptive Predictive Control of a Complicated Industrial Process

The automatic control of blast furnace (BF) ironmaking process has always been an important yet arduous task in metallurgic engineering and automation. In this article, a novel Kalman filter-based robust model-free adaptive predictive control (MFAPC) method is proposed for the direct data-driven control of molten iron quality in BF ironmaking. First, a compact-form dynamic linearization-based extended MFAPC method for multivariable molten iron quality control is proposed by generalizing the existing single-variable MFAPC method to multivariable systems. Based on it, a Kalman filter-based robust MFAPC is further proposed considering the problems of data loss and measurement noise in quality detection. Specifically, the robust mechanism in the robust MFAPC combines a novel dynamic linearization method with a concept termed Pseudo-Jacobian matrix to predict the missing data during data loss. After that, a Kalman filter is constructed based on a prediction model to filter the measurement noise. The stability of the proposed control method is analyzed, and various data experiments using actual industrial data are performed to verify the effectiveness of the proposed methods. <i>Note to Practitioners</i>&#x2014;The extremely complicated dynamics of blast furnace ironmaking process make the model-based controllers difficult to realize in practice. In this article, a novel robust model-free adaptive predictive control method is proposed for direct data-driven control of multivariate molten iron quality in the ironmaking process. This method directly uses the process input and output data to design the multivariable quality controller online by the compact-form dynamic linearization technology and the internal multilayer prediction mechanism, thus avoids the drawback of model-based controllers in troublesome process modeling. Moreover, the proposed method can effectively avoid the influence of data loss and measurement noise on the controller performance with the designed Kalman filter-based robust mechanism. The superiority and practicability of the proposed method are verified using various experiments against actual industrial data.

Output Space Transfer-Based MIMO RVFLNs Modeling for Estimation of Blast Furnace Molten Iron Quality With Missing Indexes

The final product quality of blast furnace (BF) smelting, i.e., the molten iron quality (MIQ), cannot be measured in real-time, which has become one of the important factors restricting the automation of BF. Although some data-driven soft sensing methods have been presented recently, two issues still need to be solved to meet the demand for on-site applications. The first issue is that the data-driven modeling process cannot overlook the correlation among multiple MIQ indexes (MIQIs). The second issue is that the data-driven modeling method can tolerate the missing of MIQIs, which can appear due to some factors. To address the above two issues, this article proposed an output space transfer (OST)-based multiinput-multioutput (MIMO) random vector functional-link networks (RVFLNs) for simultaneously modeling multiple MIQIs. Through sparse learning of the OST matrix, the interindicator correlation is reasonably modeled. Due to the full use of interindicator correlation, the impact of missing indicators on modeling accuracy is greatly reduced. For the corresponding knotty optimization problem, an effective alternative optimization algorithm is presented. The actual data test shows that the proposed method can greatly improve the accuracy of multiple MIQIs estimation.

A Novel Multiple-Input–Multiple-Output Random Vector Functional-Link Networks for Predicting Molten Iron Quality Indexes in Blast Furnace

The accurate prediction of molten iron quality in blast furnace (BF) can effectively improve the smelting efficiency and operation stability, which plays an important role in energy saving, emission reduction, and economic benefits improving. With the improvement of BF information acquisition technology and data storage technology, some data-driven modeling methods for simultaneously predicting multiple molten iron quality indexes have been presented. However, these modeling processes do not explore the use of the correlation information among multiple molten iron quality indexes, resulting in the suboptimal models. In addition, they do not pay attention to the molten iron quality indexes missing issue, which may lead to a great decline in modeling accuracy. To avoiding the above disadvantage, this article presents a novel multiple-input-multiple-output random vector functional-link networks. An output space transfer matrix is exploited, which not only makes full use of the correlation among all the molten iron quality indexes, but also greatly reduces the impact of index missing. For the corresponding tricky optimization problem, an alternative optimization algorithm is given, the convergence of which is guaranteed. A simulation of an actual BF data illustrates the proposed method.

Low-rank based Multi-Input Multi-Output Takagi-Sugeno fuzzy modeling for prediction of molten iron quality in blast furnace

For complex blast furnace smelting systems with large time delay, accurate prediction of molten iron quality indicators plays an important guiding role in blast furnace control. Recently, some data-driven Multi-Input Multi-Output (MIMO) modeling methods have been proposed to model multiple molten iron quality indicators including molten iron temperature (MIT), silicon content ([Si]), phosphorus content ([P]) and sulfur content ([S]). However, those data-driven MIMO models do not consider the inter-indicator correlation, which leads to the suboptimal model for the estimation of multiple molten iron quality indicators. This paper proposed a novel MIMO Takagi-Sugeno (T-S) fuzzy model with taking full account of the inter-indicator correlation. In the novel method, the inter-indicator correlation was explicitly modeled by a low-rank learning in a latent space that overcame the great challenge of jointly determining the fuzzy rules of MIMO T-S model and the inter-indicator correlation. For the corresponding optimization problem, an effective alternating optimization algorithm is presented. The validity of the proposed method is verified by simulation and comparison with some related methods on real blast furnace data.

A Novel MIMO T–S Fuzzy Modeling for Prediction of Blast Furnace Molten Iron Quality With Missing Outputs

For complex and difficult-to-control blast furnace systems with hour-level delay, accurate prediction of molten iron quality plays a very important role in guaranteeing the stable and smooth operation. Recently, some data-driven multi-input multi-output (MIMO) modeling methods have been proposed to model multiple molten iron quality indicators including molten iron temperature, silicon content ([Si]), phosphorus content ([P]), and sulfur content ([S]). However, those data-driven MIMO models ignore the interindicator correlation, which leads to the suboptimal model for the estimation of multiple molten iron quality indicators. Moreover, the above methods do not pay attention to the molten iron quality indicators missing issue, which often occurs on blast furnace. To address the above two issues, this article proposed a novel MIMO Takagi-Sugeno (T-S) fuzzy model by utilizing an output transfer matrix. In the novel method, the interindicator correlation was explicitly modeled by a low-rank learning of the correlation matrix that overcame the great challenge of jointly determining the fuzzy rules of the MIMO T-S model and the interindicator correlation. Moreover, a new complete complementary matrix can be obtained by the output transfer from the original incomplete matrix resulting from molten iron quality indicators missing issues. For the corresponding optimization problem, an effective alternating optimization algorithm is presented, and the convergence of the optimization algorithm is also rigorously proved. The validity of the proposed method is verified by comparison with some related methods on real blast furnace data.

Fault identification for quality monitoring of molten iron in blast furnace ironmaking based on KPLS with improved contribution rate

Blast furnace (BF) ironmaking is one of the most important production links in modern iron–steel making. The operation conditions of BF and the molten iron quality (MIQ) should be monitored and analyzed in real-time to realize high quality with low energy consumption. Aiming at the problems of strong nonlinearity and few fault samples in BF processes, a novel fault identification method for MIQ monitoring based on kernel partial least squares (KPLS) with improved contribution rate is proposed in this paper. First of all, a KPLS model is established with the actual historical data in order to detect the quality-related faults accurately. The T2 and SPE statistics are used to monitor the operation conditions of process from different aspects. Second, in view of the unclear physical meaning and complex computation of the existing fault identification methods based on KPLS with contribution rate, a scale factor vector is introduced into the new samples to calculate the T2 and SPE statistics, so as to construct the monitoring indicators functions. By performing Taylor approximation on these constructed functions near the scale factor with the value of 1, two new statistics are obtained from the absolute value of the first-order partial derivative representing the contribution rate of each variable. Finally, in order to further improve the effect of fault identification, the relative contribution rate of each variable is used to identify the final faulty variables. The tests of MIQ monitoring in BF ironmaking process using actual industrial data verify the validity and practicability of the proposed method.

Robust stochastic configuration network multi-output modeling of molten iron quality in blast furnace ironmaking

Blast furnace ironmaking (BFI) is currently the most widely used method of pig iron smelting. In order to achieve efficient and reasonable control, how to quickly and accurately obtain the molten iron quality (MIQ) model is a key issue. Aiming at this problem, this paper applies robust stochastic configuration networks (RSCNs) based on kernel density estimation (KDE) into the BFI modeling to obtain the MIQ model with good modeling accuracy and strong robustness quickly and effectively. Firstly, the network model is incrementally constructed by adding neurons one by one using the conventional SCNs algorithm. Secondly, in order to solve the problem of insufficient robustness of conventional SCNs, kernel density estimation algorithm is introduced to obtain the corresponding probability density estimates of each training set, and it's used as the penalty weight introduced into constructing process of conventional SCNs. At the same time, the network output weight is obtained by an improved method to solve the problem that the output weight of the conventional RSCNs is abnormal in the multi-output modeling application. Finally, modeling experiments based on actual industrial data of BFI production verified that RSCNs can achieve good modeling accuracy and strong robust performance.

Temperature Measurement and Compensation Method of Blast Furnace Molten Iron Based on Infrared Computer Vision

The temperature of molten iron at the taphole of blast furnace (BF) is an important parameter that reflects molten iron quality and BF conditions. It is not easy to measure the temperature of molten iron at the taphole in real time. To achieve continuous and accurate detection of molten iron temperature (MIT) at taphole, this paper proposes a temperature measurement and compensation method of molten iron based on infrared computer vision. First, an infrared computer vision system is designed and installed to capture the infrared thermal images of molten iron flow at the taphole. Then, the molten iron flow area is determined by using image processing. Afterward, the temperature information of slag region is obtained to calculate the MIT based on threshold segmentation. Furthermore, considering the measurement error caused by dust, the texture features influenced by dust are extracted based on the defined temperature-level co-occurrence matrix and the neighboring temperature-level-dependence matrix, and a compensation model is established to compensate the measurement error based on ensemble neural network and support vector regression. Finally, considering that the MIT and the slag temperature are approximately the same at the taphole, the MIT at the taphole is acquired according to the slag temperature. Industrial experiments and applications demonstrate that the proposed method can measure the MIT at the taphole continuously and accurately and provide reliable MIT data for operators to control BF.

Temperature Measurement Method for Blast Furnace Molten Iron Based on Infrared Thermography and Temperature Reduction Model.

The temperature measurement of blast furnace (BF) molten iron is a mandatory requirement in the ironmaking process, and the molten iron temperature is significant in estimating the molten iron quality and control blast furnace condition. However, it is not easy to realize real-time measurement of molten iron temperature because of the harsh environment in the blast furnace casthouse and the high-temperature characteristics of molten iron. To achieve continuous detection of the molten iron temperature of the blast furnace, this paper proposes a temperature measurement method based on infrared thermography and a temperature reduction model. Firstly, an infrared thermal imager is applied to capture the infrared thermal image of the molten iron flow after the skimmer. Then, based on the temperature distribution of the molten iron flow region, a temperature mapping model is established to measure the molten iron temperature after the skimmer. Finally, a temperature reduction model is developed to describe the relationship between the molten iron temperature at the taphole and skimmer, and the molten iron temperature at the taphole is calculated according to the temperature reduction model and the molten iron temperature after the skimmer. Industrial experiment results illustrate that the proposed method can achieve simultaneous measurement of molten iron temperature at the skimmer and taphole and provide reliable temperature data for regulating the blast furnace.