Prediction Of Silicon Content Research Articles

Blast furnace hot metal temperature and silicon content prediction using soft sensor based on fuzzy C-means and exogenous nonlinear autoregressive models

The temperature and silicon content of hot metal are essential parameters for the thermal control of a blast furnace. However, the physical structure of the blast furnace prevents direct and online methods from accurately predicting these parameters. In this study, we propose a new algorithm based on fuzzy c-means (FCM) and exogenous nonlinear autoregressive model (NARX) to develop a soft sensor for predicting the temperature and silicon content of hot metal. FCM is a data modeling technique that works by clustering similar data objects while separating dissimilar ones. FCM is highly effective in the identification of natural groupings in the observed data; in this case, determination of groups of operational conditions. The NARX neural network presents a model for the accurate prediction of temperature and silicon content of hot metal. The proposed algorithm was evaluated based on its efficiency in simulating the industrial process for manufacturing hot metal in a blast furnace. The results showed that a soft sensor based on FCM-NARX models has a better performance than that using the conventional NARX model.

Research on prediction classification and compensation for silicon contents in blast furnace based on ridge regression

In blast furnace, the accurate prediction of silicon contents in hot metal plays an important role in comprehensively judging the furnace states and making correct decisions. However, traditional prediction models are easily affected by multi-collinearity among the independent variables and imbalance sample distribution in hot metal datasets. The prediction values of silicon contents would be difficult to track the real values accurately when the furnace conditions fluctuate dramatically. In this paper, a classification and compensation model for silicon contents is proposed by combining the ridge regression and random forest classification algorithm. Firstly, the ridge regression algorithm is applied to fit and optimize the prediction values outputted by a front-end prediction model. Then, the random forest technology is expected to classify and identify the furnace states. Finally, the predicted values are compensated according to classification results. In this paper, the random forest regression and BP neural network algorithm were used to predict silicon content as front-end models. The related applications show that the proposed classification and compensation model improves the accuracy of predicted values effectively.

Two Derivative Algorithms of Gradient Boosting Decision Tree for Silicon Content in Blast Furnace System Prediction

The background of the present study complies with silicon content prediction in hot metal in the blast furnace system. The blast furnace system is a highly complex industrial reactor in the conventional process. The system is subject to several problems (e.g., system automation, the thermal state of the blast furnace, and the life prediction of blast furnace) that should be addressed by professionals. To determine the prediction state of the heat in the blast furnace, the silicon content in the blast furnace molten iron commonly acts as a key indicator. Based on the assumption that the blast furnace system exhibits a stable state, the accuracy of hot metal silicon is analyzed by using a range of machine learning algorithms. In the present study, two derivative algorithms of gradient boosting decision tree are adopted to develop a strong boosting predictor based on the extreme gradient boosting (XGBoost) algorithm and the light gradient boosting machine (LightGBM) algorithm for prediction. Compared with the conventional algorithms (e.g., lasso, random forest, support vector machine and gradient boosting decision tree), the prediction by using the two boosting algorithms is capable of more effectively guiding and determining the state of the blast furnace. As revealed from experimentally simulated results, the mentioned two boosting algorithms exhibit better comprehensive prediction performance than the conventional algorithms on the datasets of two practical blast furnace systems, demonstrating that the R-square of the two blast furnaces in the training set is over 0.7. The mentioned two algorithms are of certain guiding significance for exploring blast furnace problems.

Prediction method of silicon content in blast furnace hot metal based on IPSO-HKELM

Aiming at the problem that the silicon content of molten iron can not be detected online, a model for predicting silicon content in molten iron based on Hybrid Kernel Extreme Learning Machine optimized by Improved Particle Swarm Optimization Algorithm (IPSO-HKELM) is proposed. Firstly, the input variables are reduced by PCA, and then the prediction model of molten iron content based on HKELM is established. In this paper, PSO is used to optimize the kernel parameters of HKELM. Aiming at the problem that PSO is easy to fall into local optimum, the Inertia weight reduced with the number of iterations and the random back-based learning mutation operation are introduced, so that PSO can jump out of the local minimum point more easily and get the optimal result. Experiments show that the prediction model of silicon-based silicon content based on IPSO-HKELM has high prediction accuracy and short time, which can meet the actual production needs.

Soft Sensing of Silicon Content via Bagging Local Semi-Supervised Models.

The silicon content in industrial blast furnaces is difficult to measure directly online. Traditional soft sensors do not efficiently utilize useful information hidden in process variables. In this work, bagging local semi-supervised models (BLSM) for online silicon content prediction are proposed. They integrate the bagging strategy, the just-in-time-learning manner, and the semi-supervised extreme learning machine into a unified soft sensing framework. With the online semi-supervised learning method, the valuable information hidden in unlabeled data can be explored and absorbed into the prediction model. The application results to an industrial blast furnace show that BLSM has better prediction performance compared with other supervised soft sensors.

Application of grey relational analysis and extreme learning machine method for predicting silicon content of molten iron in blast furnace

ABSTRACTControlling the molten iron temperature plays an important role in the iron and steelmaking industry. The change of silicon content is adopted to reflect the temperature, however , the prediction of silicon content has been one of the hot and difficult problems. In this paper, a new model based on gray relational analysis (GRA) and extreme learning machine (ELM) is developed. Firstly, the GRA is used to get the high correlation indexes with the silicon content. Then the relevant indicators are taken as input and the silicon content is taken as output. The ELM model is constructed and the model is trained. Based on this, the silicon content is predicted. The results show that the hit rate reaches 87%(the error is less than 0.10). Compared with the traditional backpropagation or radial basis function neural network , this model has higher hit rate and faster running speed.

Introduction into MMK Blast Furnaces of an Automated System for Smelting Control, Optimization, and Prediction

Results are provided for work conducted in the Magnitogorsk Metallurgical Combine blast-furnace workshop for production satisfaction of a blast-furnace automatic control system by stagewise introduction of local systems combining into one automated system for control, optimization, and prediction of blast furnace smelting (ASCOP BF). Using mathematical models and special ASCOP BF algorithms for each blast furnace, there is continuous automatic monitoring of radial and circumferential gas distribution in the bell, circumferential gas distribution in the furnace hearth, material and thermal balances of smelting and furnace heating; optimization of the smelting gas dynamic regime, radial and circumferential gas distribution in the bell, circumferential gas distribution in the furnace hearth, smelting recovery processes, thermal state of a furnace and melting zone parameters; prediction of silicon content in cast iron. The order of conducting operations is proposed and substantiated for achieving maximum productivity and minimum coke consumption with given smelting raw material and operating conditions. It is shown that the mathematical models used in the system by the process described are ready for use in blast furnace automatic control systems already created for blast furnaces in other metallurgical enterprises in order to achieve the best smelting indices.

Just-in-Time Correntropy Soft Sensor with Noisy Data for Industrial Silicon Content Prediction

Development of accurate data-driven quality prediction models for industrial blast furnaces encounters several challenges mainly because the collected data are nonlinear, non-Gaussian, and uneven distributed. A just-in-time correntropy-based local soft sensing approach is presented to predict the silicon content in this work. Without cumbersome efforts for outlier detection, a correntropy support vector regression (CSVR) modeling framework is proposed to deal with the soft sensor development and outlier detection simultaneously. Moreover, with a continuous updating database and a clustering strategy, a just-in-time CSVR (JCSVR) method is developed. Consequently, more accurate prediction and efficient implementations of JCSVR can be achieved. Better prediction performance of JCSVR is validated on the online silicon content prediction, compared with traditional soft sensors.

A Sliding-Window T-S Fuzzy Neural Network Model for Prediction of Silicon Content in Hot Metal

Iron making in blast furnace is one of the most complicated industrial processes, especially in its dynamics, inertial properties and multi-scale availabilities. Over the years, researchers have been using silicon content to judge the temperature and the conditions within the blast furnace due to the complexity in measuring the actual status that results from extreme temperatures and intricate environment. Addressing these limitations, a sliding-window Takagi-Sugeno fuzzy neural network(SW-TS FNN) model is proposed to predict the silicon content in hot metal. Through the sliding of a proper width of the sliding-window, the train data for T-S fuzzy neural network(FNN) model can be updated at desired time increments, giving the latest prediction of silicon content. Compared to a simple T-S FNN model on the prediction of silicon content, this SW-TS FNN model shows great improvement at hit rate and mean-square error.

Prediction of the hot metal silicon content in blast furnace based on extreme learning machine

Silicon content in hot metal is an important indicator for the thermal condition inside the blast furnace in the iron-making process. The operators often refer the silicon content and its change trend for the guidance of next production. In this paper, we establish the neural network model for the prediction of silicon content in hot metal based on extreme learning machine (ELM) algorithm. Considering the imbalanced operating data, weighted ELM (W-ELM) algorithm is employed to make prediction for the change trend of silicon content. The outliers hidden in the real production data often tend to undermine the accuracy of prediction model. First, an outlier detection method based on W-ELM model is proposed from a statistical view. Then we modified the ordinary ELM and W-ELM algorithms in order to reduce the interference of outliers, and proposed two enhanced ELM frameworks respectively for regression and classification applications. In the simulation part, the real operating data is employed to verify the better performance of the proposed algorithm.

Ensemble Non-Gaussian Local Regression for Industrial Silicon Content Prediction

Ensemble Non-Gaussian Local Regression for Industrial Silicon Content Prediction

Silicon content prediction and industrial analysis on blast furnace using support vector regression combined with clustering algorithms

Silicon content prediction is quite significant for supervising the state of blast furnace and is usually selected as the indicator to represent the thermal state. In practical industry, the fluctuation exists in the operation of blast furnace all the time. What's worse, it is inaccurate to build the predictive model with many outliers. To solve these problems, this paper has developed a model to predict the silicon content using support vector regression (SVR) combined with clustering algorithms, including hard C-means (HCM) clustering and fuzzy C-means (FCM) clustering. Through data processing, the data points are clustered based on the similarity, and then different SVR models are established. In order to make full use of FCM, a new method using multiple SVRs and FCM based on membership degree (MFCM-SVRs) is proposed where the membership degree is applied to eliminate the outliers. Simulation results verify that the multiple SVRs based on HCM (HCM-SVRs) and MFCM-SVRs possess superiority in terms of accuracy and speed, which makes the method serve better for practical production.

A modified ELM algorithm for the prediction of silicon content in hot metal

Silicon content in hot metal in the iron-making process has been used as a major indicator of the internal thermal state in blast furnace for many years. Due to the harsh environment in the blast furnace, how to measure the silicon content in hot metal is quite difficult. Many efforts have been made on the estimation of the silicon content in hot metal. In this paper, a soft-sensing modeling method based on a modified extreme learning machine is proposed to tackle the problem. In this approach, a modified pruning algorithm is utilized to optimize the weights which are generated randomly in the original ELM algorithm. The real data collected from a blast furnace in the factory are applied and tested by the proposed algorithm, and the results show that the proposed prediction model has less error than the other algorithm such as BP algorithm and support vector method.

Data-Driven Time Discrete Models for Dynamic Prediction of the Hot Metal Silicon Content in the Blast Furnace—A Review

A review of black-box models for short-term time-discrete prediction of the silicon content of hot metal produced in blast furnaces is presented. The review is primarily focused on work presented in journal papers, but still includes some early conference papers (published before 1990) which have a clear contribution to the field. Linear and nonlinear models are treated separately, and within each group a rough subdivision according to the model type is made. Within each subsection the models are treated (almost) chronologically, presenting the principle behind the modeling approach, the signals used and the main findings in terms of accuracy and usefulness. Finally, in the final section the approaches are discussed and some potential lines of future research are proposed. In an Appendix , a list of commonly used input and output variables in the models is presented.

Prediction of Silicon Content in Hot Metal Based On Integrated Neural Network

Prediction of Silicon Content in Hot Metal Based On Integrated Neural Network

Bayesian Regularization Neural Networks For Prediction Of Silicon Content In Blast Furnace

Bayesian Regularization Neural Networks For Prediction Of Silicon Content In Blast Furnace

Applying Fuzzy C-means Clustering and Multiple SVM to Silicon Content Prediction in Hot Metal

Applying Fuzzy C-means Clustering and Multiple SVM to Silicon Content Prediction in Hot Metal

Improving the Solution of Least Squares Support Vector Machines with Application to a Blast Furnace System

The solution of least squares support vector machines (LS‐SVMs) is characterized by a specific linear system, that is, a saddle point system. Approaches for its numerical solutions such as conjugate methods Sykens and Vandewalle (1999) and null space methods Chu et al. (2005) have been proposed. To speed up the solution of LS‐SVM, this paper employs the minimal residual (MINRES) method to solve the above saddle point system directly. Theoretical analysis indicates that the MINRES method is more efficient than the conjugate gradient method and the null space method for solving the saddle point system. Experiments on benchmark data sets show that compared with mainstream algorithms for LS‐SVM, the proposed approach significantly reduces the training time and keeps comparable accuracy. To heel, the LS‐SVM based on MINRES method is used to track a practical problem originated from blast furnace iron‐making process: changing trend prediction of silicon content in hot metal. The MINRES method‐based LS‐SVM can effectively perform feature reduction and model selection simultaneously, so it is a practical tool for the silicon trend prediction task.

A Predictive Dynamic Neural Network Model Based on Principal Component Analysis (PCA) and its Application

We propose a real-time prediction method based on PCA and improved multi-step Elman net. The method not only preserves most original data information, but also eliminates the relativity among data and simplifies the net structure. It can predict the complex and nonlinear systems with dynamic recurrent algorithm. Through training the net has the ability of adapting to the uncertainty of the nonlinear structure, and then reflects the dynamic character of the systems. The hit rate reaches 88.17% to forecast the silicon content in hot metal of a blast furnace with errors ranging from-0.05 percent to 0.05 percent. The results prove that this method is feasible to predict silicon content on blast furnace and also it’s a prediction method of nice future.

Silicon Content Prediction Using the Hybrid Model by Fuzzy C-means Clustering and Artificial Neural Networks

Silicon Content Prediction Using the Hybrid Model by Fuzzy C-means Clustering and Artificial Neural Networks