Identification of superheat of aluminum electrolytic cell based on computer vision and expert rule

In the aluminum electrolysis production process, the degree of superheat is an influential index of the whole process, reflecting the distribution of the physical field of the electrolytic cell, the current efficiency in the electrolytic production, and the lifespan of the electrolytic cell. Therefore, detection or identification of the superheat is significant for the production process. However, as the temperature detection of the primary crystal needs a long period off-line experiment, real-time measurement of superheat is still a problem beyond solution. Considering the strong link between the fire hole observation and superheat degree, a method of superheat degree identification based on computer vision technology and expert rules is proposed in this paper. First, the corner detection and the optical flow method are used to extract the fluctuation amplitude features from the flame hole video. Then, the color change feature is extracted according to the variation of the brightness value of the fire area of the sampling frame. Finally, some expert rules are developed to determine the degree of superheat. In this method, the computer vision technology is applied to superheat identification for the first time with valid verification by experimental results of superheat degree.