The BOF Endpoint Control Model And Intelligent Expert System Based On Neural Network

The steel industry is one of the most important industries of our country, and the BOFoccupies an extremely important position in the steel industry. The main task is to control thecarbon content and the tapping temperature of the endpoint, so that it can meet the standardsteel. However the BOF steelmaking is a very complicated industrial procedure that includedmany physical and chemical reactions. It is also a periodic procedure of temperature rise andcarbon drop. In addition,it is very difficult to detect the steel content and temperature duringthe procedure because of the converter temperature is extremely high.As a result,it is hardly tocontrol it by conventional measure.Currently,most of the small and medium converters arecontrolled by a subjectivity manual operation mode which based on the experience ofworkers.The control accuracy of this mode is too low to get a desired endpoint that will oftenneed a rework and bring huge economic loss.Therefore,the BOF endpoint precise control isimportant.The intelligent control brings a new way to the BOF endpoint control.Neural networkis an important part of the intelligent control which has a strong ability of solving nonlinearsystems and has been widely used in the BOF endpoint control.In addition,the theoreticalmodel is the basis for steelmaking endpoint control which has a huge impact on the endpointprediction.As a result,combining the theoretical model with neural network is a valuableresearch way.The paper firstly elaborated the static control model theoretical models and figure outthe initial parameters through a detailed analysis of BOF steelmaking process. On the basis ofthe parameters obtained and then used the multiple neural networks and case-based reasoningexpert system to predict those sample data, the pedigree cluster divide the large data sets intoseveral categories,the degree of similarity will be relatively high in each category afterdivision,then train neural model for every category.Finally make predictions.Simulationresults show that the multi-neural network model has better prediction results.The case-basedreasoning expert system has been used for the abnormal samples to make the prediction morecomprehensive.This paper secondly introduces the BOF endpoint control model and software,animportant application in the control model,the ant colony algorithm for the BOF endpointcontrolled multi-objective optimization.The main part of the software developed by VisualStudio2005, the back ground using the SQL Server database,its main function is to predictthe end of steelmaking and control，all the work done by the end of the text of a summary andgeneralization,and the next stage of work needs to be done on the basis of the outlook.