Font Size: a A A

Research On Ladle Jacking Height Prediction Based On Industrial Camera

Posted on:2022-12-02Degree:MasterType:Thesis
Country:ChinaCandidate:X Y ZhangFull Text:PDF
GTID:2481306779463044Subject:Computer Software and Application of Computer
Abstract/Summary:PDF Full Text Request
In the production and refining process of RH steelmaking process,it is necessary to move and jack the ladle containing molten iron.The ladle needs to be lifted to a high place to make the upper impregnation pipe go deep below the liquid steel level in the ladle for degassing reaction.The lifting height is directly related to the distance between the liquid steel and the opening of the ladle and the thickness of the slag layer.At present,this step is still observed by experienced steelmakers through naked eyes and manually control the ladle to the appropriate position.Because the liquid steel with temperature up to 1650 ? radiates very much,it does great damage to people's eyes,and there are huge potential safety hazards to the stability and reliability of jacking.There is an urgent need for automatic methods including automatic measurement and automatic calculation of jacking height to realize the automation of ladle jacking.In industrial production,the traditional size detection mostly adopts proximity or infrared ranging equipment.For the high-temperature target in the scene of steel refining,the method of measuring liquid steel level and slag layer thickness by proximity sensor is obviously not applicable.Infrared ranging equipment will also cause large measurement errors due to the high temperature and radiation on site.The measurement method using machine vision can be separated from the target to be measured for information collection and processing.In this paper,the industrial camera is used to collect and process the equipment and the machine vision algorithm is used.The main research contents and work are as follows:(1)Selection of acquisition equipment: according to the temperature and radiation conditions in the on-site environment of RH refining workshop and the characteristics of the target to be measured,combined with the comprehensive consideration of different image sensors.In this paper,MER2-302-37G/C Daheng industrial camera and conical high temperature resistant lens are selected as acquisition equipment to take images of the target ladle and impregnation tube.(2)RH model shooting scheme planning: through the camera imaging principle,the imaging models of liquid steel ladle and immersion tube are constructed respectively,the mapping relationship between target pixel distance and real distance in the image is deduced,and the transformation formula of pixel distance real distance is obtained.(3)Calculation of distance between liquid steel surface and ladle opening: considering that the ladle containing liquid steel in the steelmaking plant is a kind of cylinder,the image information is observed and collected from the top side view of the ladle,and the target area is oval.The ellipse fitting algorithm based on arc clustering is used to obtain the parameter information of the ellipse model of the liquid steel level and the ladle opening,calculate the pixel height difference of the upper vertex in the short half axis direction,and use the transformation formula of the pixel distance to obtain the estimated value of the real distance.(4)Calculation of steel slag layer thickness: considering that the impregnation pipe separated from liquid steel has two parts in contact with steel slag and liquid steel,they will leave delamination information on the outside of the impregnation pipe.Collect the top view angle image information of the impregnation tube side,segment and extract the steel slag contact area outside the impregnation tube through the watershed algorithm guided by distance transformation,obtain the pixel height difference of the steel slag layer according to the coordinate information of its upper and lower edges,and then obtain the estimated value of the real thickness by using the transformation formula of pixel distance.Based on the comprehensive safety distance,the ladle jacking formula is obtained.The lifting distance of ladle is estimated according to the actual image,so as to reduce human eye observation and manual control,and realize the automation of RH refining ladle lifting.
Keywords/Search Tags:Machine vision, RH refining method, Ellipse detection, Image segmentation, Watershed algorithm
PDF Full Text Request
Related items