Artificial Neural Network Model for Predict of Silicon Content in Hot Metal Blast Furnace

Abstract

The growing focus on the efficiency of the reduction process in blast furnace generates an alteration in the way they operate. This modifies the conditions of transfer of silicon for the hot metal and can cause problems in the added value of your product. To evaluate the changes of the operational parameters of the reduction on the conditions of transfer of silicon process a mathematical model based on artificial neural networks has been implemented. Through this model it was possible to predict the silicon content to determine the influence of each operational parameter. Artificial neural networks were able to predict the silicon content through parameters of the reduction in blast furnace process, and this was verified by the precision of this model. The ANN showed that Theoretical flame temperature, Pressure blow and Coke rate have a positive influence on the silicon content in hot metal, and the Hot metal rate is inversely proportional to the silicon content of the hot metal.

Artificial Neural Network for Predicting Silicon Content in the Hot Metal Produced in a Blast Furnace Fueled by Metallurgical Coke

“…Several authors argue that normally 85% of the data is used to train and validate the neural network and the remaining 15% of the dataset is presented to the neural network only when its performance in the training and validation phase is considered satisfactory 2,16,30 .…”

“…The test step is important to evaluate the generalization and learning ability of the neural network. The neural network must be able to generalize what it has learned and reproduce the solutions from the trained examples for any problems similar to the training 2,16 .…”

“…High values of silicon content indicate increased heat input to the furnace and in some cases may indicate excess coke in the furnace. Since coke costs predominate in the production of cast iron, tighter control of silicon content therefore clearly has economic advantages 7,[15][16][17][18][19] .…”

“…According to the literature, several authors suggest that one or two layers are sufficient for modeling metallurgical problems. Only one hidden layer, using a sigmoid-type activation function, is sufficient for the network to converge to good results as long as there are enough input variables to train the algorithm 2,16,28 .…”

Static Heat Energy Balance Mathematical Model for an Iron Blast Furnace

“…Taking [1] as reference, the following study has been conducted to calculate exact heat distributions of each input and output component of the furnace, in detail. Various standard textbooks [2][3][4][5], and papers [11][12][13][14] have been referred to get an in-depth insight into the factors which should be taken into consideration, and the reactions taking place in the different regions of the furnace. The standard enthalpy of formation of compounds data has been taken from [6], the enthalpy data of elements and compounds at high temperatures has been taken from [7] and the heat of solution data has been taken from [8].…”