Research On Development And Application Of Intelligent Warning Terminal For Pre-ironmaking Blast Furnace

Blast furnace smelting occupies an important position in the entire chain of the iron and steel industry,but the internal smelting process of the blast furnace is complicated,so accurate and effective evaluation of blast furnace operation status has become a research hotspot in recent years.In order to ensure that the blast furnace produces high-quality molten iron,the blast furnace operator needs to take measures to control it in advance,so it is of great practical significance to accurately predict the quality of the molten iron before tapping.The quality prediction of molten iron and how to monitor the operating status of blast furnace in real time are investigated by this project.The work mainly includes the following two parts:(1)The quality of molten iron is mainly measured by silicon and sulfur components.A novel prediction model based on the gated recurrent unit neural network GRU is proposed in order to solve the problem of insufficient prediction efficiency and accuracy of silicon and sulfur content in molten iron,The GRU neural network can solve the nonlinear problem well and fully mines the time sequential relationship in the actual production data of the blast furnace.Raw materials,blast furnace smelting process data and control parameters are used as model input variables,molten iron silicon and sulfur content are used as output variables,and Pearson correlation analysis and cross correlation analysis methods are adopted by this thesis.The hysteresis of the influence of blast furnace input variables on the quality of molten iron are analyzed firstly,then the GRU model is adopted to test on the pre-processed data set,combing GRU with linear regression model LR,multiple feedback neural network model BP,long short-term memory neural network model LSTM to compare with the forecast results.Experimental results show that the GRU model is more accurate in predicting the silicon and sulfur content of molten iron in blast furnaces.This method provides a new way for the prediction of molten iron quality in blast furnaces.With the help of accurate molten iron quality prediction,the blast furnace operators can be guided to operate in advance to avoid affecting the molten iron quality.(2)The intelligent warning terminal for the iron front blast furnace was designed and implemented combined with the molten iron quality prediction model,The terminal not only realizes the real-time prediction of the molten iron quality of the blast furnace,but also develops the online monitoring of key operating parameters of the blast furnace and the warning push function of over-limit,so that the blast furnace operator can view the current operation of the blast furnace anytime and anywhere,guide production in time,and ensure the smooth run of the blast furnace.