Research And Application Of Blast Furnace Material Distribution Decision System Based On Data Driven And Mechanism Analysis

As the upstream process of iron and steel metallurgical industry,blast furnace ironmaking is the main link of CO2 emission and energy consumption in iron and steel industry.With the continuous development of computer storage technology,operation ability,and improve the sensor performance brought by blast furnace internal detection parameters is more and more abundant,accurate,large number of internal operation indicators to blast furnace detection and long-term storage,data provide the basis for the establishment of blast furnace operation driven model.In this paper,the machine learning and data mining technology are widely used in machine learning.The control system of blast furnace burden distribution is studied and developed,and the operation status of blast furnace is analyzed in real time.Based on the analysis of the operating mechanism of the blast furnace,the correlation between the top parameters is obtained,and then the data mining algorithm is used to excavate the potential correlation between the parameters.Considering the time delay characteristics of the blast furnace system,get the relevant parameters of different lag time from the perspective of statistical analysis,establish the operation condition of blast furnace data driven model to estimate the change trend of quasi important index of top temperature,pressure etc..The prediction model not only enrich the existing observable data,and provide more information for the model auxiliary material control method;blast furnace temperature,pressure,gas flow parameters such as furnace measure is furnace furnace due to the lack of historical data information standard,so need to use unsupervised learning technique.The clustering analysis of blast furnace operation model set.Then,according to the current parameters of the furnace roof and the matching of the multi model set,the furnace condition is judged.Then the reasonable distribution system is calculated for each furnace condition.Based on the traditional surface shape model,put forward the model of Gauss function and trigonometric function profile based on contour surface fitting can not only better,and the parameters in the model are easier to identify,suitable for computer programming and calculation process model of fabric.Finally,using DEM simulation method to verify the model,DEM to replace the traditional cold blast furnace experiment method,not only can save manpower and material resources,but also can analyze the stress on burden,find the material collapse,mechanical principle leduced the abnormal situation.Finally,the prediction algorithm,multi model control algorithm programming,based on the C# programming language,SQL server database technology,MATLAB simulation software development of the control system of blast furnace cloth.First of all,the MATLAB algorithm is used for off-line verification,and then the C# language is used to read the data from the background program,and the important index parameters are predicted and classified.Finally,it is displayed in the interface system developed by C#.Enrich the visualization of the top data,and give the suggestion of the cloth system based on the data analysis.