An integrating system for predicting Si content in pig iron of blast furnaces

In this paper, an integrating system for predicting Si content in pig iron of blast furnaces is presented which integrates a neural network model with qualitative analysis. Through causal analysis and qualitative reasoning, the qualitative trend of the process in blast furnace is predicted, and the relevant variables and model structure are determined. Then, a neural network model is constructed and trained with appropriate data automatically. With the model, Si content in pig iron will be predicted. Evaluation of the system was made by comparing the predicting values with observed values, and excellent results were achieved