Modeling And Optimization Of Ladle Furnace Alloy Feeding System

The ladle furnace is one of the most important refining units. The alloy composition adjustment is one of its main functions. Therefore, the alloy feeding system has become the important research object in the ladle furnace research area. However, the existing adjust scheme and alloy burden model are not satisfied, in the actual product operation alloy composition adjustment often bases on experience. It does not only affect the quality of molten steel, but also does not optimize the alloy feeding system reasonably. Therefore, in this thesis the research content is modeling and optimization of the ladle furnace alloy feeding system.The ladle furnace alloy feeding process requires that once feeding operation can make the alloy composition of molten steel to meet the requirement of the steel grade, and insure the cost of alloy materials is lowest. For the former, it needs to get the alloy elements yield; for the latter it needs to establish the alloy burden optimal model. The element yield is an important parameter of alloy burden optimal model, and it is also a great influencing factor to the accuracy of solutions. But the alloy elements yield are very difficult to obtain, therefore, in this thesis it first selects the variable factors which influence the alloy elements yield through the metallurgical mechanism analysis, and then establish the incremental prediction model of case-based reasoning and RBF neural network to forecast the alloy elements yield. By using the solution of the optimal model, the optimal scheme of the alloy feeding process is obtained. The alloy elements yield prediction model and the alloy burden optimal model constitute the optimization model of alloy feeding system together.Through the simulation of the actual production data, it is proved the method in this thesis is effective. And it has guiding meaningful to the actual industrial production.