Research On Alloying Component Controlling Model For Lf Steelmaking Based On Hs-svr Algorithm

Ladle furnace has become one of the key equipments in secondary refining process. It plays an important role in the control of components of molten steel, especially the alloy components. Now, the research on alloying component controlling has become one of the main subjects in ladle furnace research area. But, for a long time, the existing controlling model doesn't satisfy the requirement. It is generally based on the experience of operators to adjust alloy components. This does not only affect the control accuracy, but also obstruct getting an optimal solution of alloy compounding. Therefore, in this thesis, an intensive study of alloying component controlling model of ladle furnace has been conducted.In order to achieve optimal compounding, A minimum cost for the objective of linear programming is established and the minimal solution can be gotten. But, the model's precision greatly depends on the element yield which is very difficult to be obtained. Followed by analyzing the factors which influence the element yield, a prediction model for element yield is established using GA & SVM algorithm. As new samples are growing, the training set of element yield rate model is also becoming larger. To reduce the prediction time and improve predicting accuracy, an incremental learning algorithm, HS-SVR, is proposed. This incremental learning algorithm can reduce the size of training set of prediction model. An element yield rate can be predicted using prediction model with HS-SVR algorithm. Using optimal alloy compounding model with the predicted value, an optimal feeding solution can be obtained at last. The alloying component controlling model consists of prediction model of element yield rate, incremental learning algorithm and optimal compounding model.The experiments using LF furnace data, show that, the model can guarantee that in accordance with the requirements of steel components, it can achieve accurate, economic and quickly.alloying operation. Narrow component controlling will be achieved at last.