A predicting system based on combining an adaptive predictor and a knowledge base as applied to a blast furnace

In this paper a system is proposed and applied to predicting silicon content in pig iron for a blast furnace. It combines an adaptive predictor with a knowledge base. On-line experiments show that the prediction accuracy of the system is much better than that of an experienced operator or of an adaptive predictor alone. Such a system will be helpful to operators in selecting desirable control methods.