Modeling And Control Of Gas Flow In Blast Furnace Smelting Process

Blast furnace smelting is an important link in the production of iron and steel.Whether the blast furnace condition can be stable and smooth for a long time is directly related to the quality of pig iron,smelting cost and gas emission,etc.In the process of blast furnace,due to the complex reaction in the blast furnace,especially the characteristics of large lag,it is difficult to judge the blast furnace.The blast furnace gas flow has the characteristics of real-time and rich information of blast furnace,and it becomes an important information to reflect the situation of blast furnace.Therefore,this paper combines the knowledge of the blast furnace smelting and the data of the blast furnace and the expert knowledge of the blast furnace establish prediction model of blast furnace gas flow.The model has been tested on the 2# BF.The model can predict the change trend of the blast furnace gas flow accurately,and then provide the help for the blast furnace condition analysis,guide the operation of the blast furnace,and ensure the smooth operation of the blast furnace.Therefore,this paper not only has a certain theoretical value,but also has a higher practical value.The main research work of this paper is given as follows:First of all,from the aspects of the mechanism of blast furnace blast furnace gas flow and the formation process of gas flow in blast furnace internal movement and change of gas flow components,and analyzes the relationship between the gas flow and blast furnace gas flow and distribution of several typical development type,including edge blowing pipe type and stroke Center.Secondly,the paper analyzes the three distribution of blast furnace gas flow in the blast furnace,and gets the output variables of the model.From the point of view of data driven,using the clustering method to cluster the input data of blast furnace,the blast furnace smelting process is divided into different furnace conditions.The improved sparse subspace clustering method,after adding the appropriate regularization term in the objective function,further processes the noise points and outliers in the blast furnace data,so that the accuracy of the prediction model is more accurate.Further,after the blast furnace data are clustered,the output variables of the model are predicted by support vector regression.Support vector regression can avoid overfitting effectively,and is more suitable for industrial complex system modeling than other intelligent algorithms.Finally,put forward the control method of output variables of the model,and the model and application of 2# BF on debugging and validation.The results show that the model can accurately predict the trend of the blast furnace gas flow,thus providing some guidance and help for the analysis of blast furnace conditions,and ensuring the smooth operation of the blast furnace.