Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
10151202 | Neurocomputing | 2018 | 7 Pages |
Abstract
Slag crusts on the cooling stave of a blast furnace offer greatly protection for the furnace wall. Frequent forming and shedding of slag crusts (FSSCs) cause severe erosion on the wall, which affects the life of a blast furnace. However, it is very difficult to detect the FSSCs directly because of restrictions imposed by the structure of a blast furnace and detecting costs. This paper presents a soft-sensing method based on two-dimensional decision fusion to detect the state of slag crust (SSC) of a blast furnace. First, a soft-sensing scheme for SSC is put forward on the basis of features of slag crusts and the temperature detected in cooling stave. Next, methods for calculating a temperature threshold (TT) and a change-rate threshold of temperature (CRTT) are presented according to the characteristics of slag crusts. Finally, a two-dimensional decision method is presented by fusing the TT and CRTT to determine the SSC of the blast furnace. The experiment results based on industrial data demonstrate the effectiveness of the method.
Keywords
Related Topics
Physical Sciences and Engineering
Computer Science
Artificial Intelligence
Authors
Jianqi An, Jialiang Zhang, Min Wu, Jinhua She, Takao Terano,