Establish the multi-source data fusion model of the shape of blast furnace burden surface based on co-universal kriging estimation method

摘要

This paper presents a multi-source data fusion model method which could improve the blast furnace (BF) burden surface model accuracy. First, the three sections of straight line are used to describe the cross section of BF burden surface, and apply the motion law of the furnace burden to constrain the specific parameters of the three sections of straight line. Secondly, a multi-source data fusion method based on co-universal kriging estimation method is proposed. The temperature and height data are combined to build the unbiased estimation for the burden surface shape. Finally, an example of surface shape model using our proposed method in a 2500 m(3) BF of a steel plant is discussed. The application shows that, contrasted with the traditional model, the model accuracy has arisen by 8%, and the resolution of surface shape has arisen by 0.32. The novel method can provide necessary guidance for energy saving and emission reduction in operation of the BF.