A visual slag peeling detection method in blast furnace

Abstract

Slag peeling is one of the important factors affecting the stable operation of blast furnaces. On the one hand, it causes a sharp rise and fall in the temperature of the furnace wall surface, resulting in significant thermal shock and erosion of the furnace wall and cooling stave. On the other hand, the temperature of the molten iron decreases when the slag falls into the hearth. If not treated in time, waste products may be generated due to high sulphur content in the molten iron, and even lead to severe furnace cooling. In order to solve this issue, this article proposes a rapid slag peeling detection method based on image processing, which contributes to the stable operation of the blast furnace. First of all, the raceway image acquired by CCD camera is greyed and denoised to speed up the detection and improve the image quality. Second, frame difference method is used to construct background template for raceway video sequence, which is the key step of slag peeling detection. In addition, considering the change of background template caused by the fluctuation of pulverised coal injection, an adaptive updating method of background template is proposed. Finally, the difference between the video sequence and the background template is applied to determine whether there is slag-peeling phenomenon. Massive raceway videos collected from a 2500 [Formula: see text] are used to evaluate the proposed detection method. The experiment results show that the method can effectively achieve slag peeling detection.