An Adaptive System With Knowledge Base for Predicting Silicon Content in Pig Iron
带知识库的高炉铁水含硅量的自适应预报系统

A self-tuning predictor and a knowledge based system are used in collaboration for predicting silicon content in pig iron for a blast furnace. It is constructed on the base of Box and Jenkin's principle. Model parameters are estimated by means of a recursive approximate maximum likelihood method. Prediction accuracy obtained by this system is much better than that obtained by an experienced operator. It can be used as an operator's guide in selecting appropriate control actions.