Application Of Machine Vision Technology In Intelligent Steelmaking

With the continuous development of steel production in China and the continuous improvement of production technology,it has a greater dependence on the production experience of the operators.At the same time,a series of factors such as the working habits of the operators affect the efficiency and optimization control of steelmaking,so there is room for further improvement.In the traditional process,sampling,temperature identification by naked eyes,carbon estimation by steel bloom,oxygen and sulfur estimation by slag inspection and other operations further increase the requirements of product quality for production personnel to rely on experience to operate accurately.Therefore,it is necessary to informationize the process parameters,digitize the raw material status,develop intelligent manufacturing,and establish intelligent and automatic modern steel-making mode,so as to reduce manual operation,reduce production risk,and ensure safety To ensure production quality and realize intelligent,energy-saving and emission reduction steel-making.In this paper,the application of machine vision technology in intelligent steelmaking is taken as the research goal.The real-time judgment of the level of molten steel churning in LF(ladle furnace)furnace and the real-time monitoring of slab surface temperature in secondary cooling zone of continuous casting are realized in intelligent steelmaking control.The recognition system of molten steel churning level in LF furnace and the monitoring system of slab surface temperature in secondary cooling zone of continuous casting are developed.Argon is injected from the bottom of molten steel in LF furnace at a predetermined flow rate.Good argon stirring is conducive to the chemical reaction between steel and slag in LF furnace,can accelerate the material transfer between steel and slag,and make the temperature of molten steel tend to be uniform.However,in the actual production process,the argon nozzle is likely to be blocked.According to the setting of argon flow rate,the real-time situation of molten steel churning can not be completely mastered,which affects the quality of output.The recognition system of molten steel level churning in LF furnace uses high-definition video camera to obtain the video information of molten steel in furnace.Based on the statistical characteristics of a single image obtained by statistical histogram,the image is preprocessed.Based on the difference of gray changes between slag and molten steel in different molten steel churning States,combined with the inter frame difference method of moving target detection,the judgment of molten steel churning level is realized.The rapid development of continuous casting technology is an important part of the development of contemporary iron and steel industry.Good secondary cooling control is the guarantee of continuous caster production capacity and slab quality requirements,and the monitoring of slab surface temperature is the basic requirement of secondary cooling control.During the cooling of continuous casting slab in the secondary cooling zone of continuous casting,the method is to spray water mist directly on the slab to reduce the surface temperature of continuous casting slab through heat absorption and evaporation of water mist.However,Due to the existence of water mist,the external infrared radiation of continuous casting slab is irregularly attenuated,which causes a noticeable interference to the non-contact surface temperature monitoring of continuous casting slab in the secondary cooling zone.The surface temperature monitoring system of continuous casting slab in secondary cooling zone adopts infrared temperature measuring device.Based on Mie scattering theory,the interference of water mist in secondary cooling zone on infrared temperature measurement is corrected by processing and calculating image information.The data exchange between detection system and control system adopts Oracle database management system.Through the Oracle database client,the results of molten steel churning and slab surface temperature in the secondary cooling zone are sent to the Oracle database server for real-time feedback to the intelligent steelmaking control.The control system can react quickly and accurately according to the feedback results,reducing the dependence on manual operation.