| With the continuous advancement of the intelligent manufacturing technology in the traditional manufacturing industry,in the key stage of the metallurgical industry's upgrading and transformation,further improving the level of automation and intelligence is an urgent problem for the industry.Equipping robots with eyes and applying machine vision to the robot system is a breakthrough in the intelligent upgrade of the industry.The galvanizing production line is a harsh production scene with high temperature,high dust,and high noise.The operators are labor intensive,and there are serious safety hazards on site.Aiming at the production environment,slag removal process and functional requirements of Handan Iron and Steel 2# galvanizing line,this paper designed and developed a set of machine vision-based slag removal robot system,and finally successfully put it on line,innovatively applying machine vision technology to plating The zinc production line has made up for the shortcomings of the first generation of teaching-based slag removal robots,greatly improved the flexibility of robot movements and the efficiency of slag removal,reduced dependence on labor,liberated labor,realized downsizing and increasing efficiency,and improved the follow-up The application of the robot vision system in the metallurgical industry has valuable reference significance.The main contents of this paper are as follows:1.According to the slag cleaning process and functional requirements of the galvanizing line,developing a slag cleaning robot system that integrates machine vision.In terms of hardware selection,Fussen Well WBOX3401 is selected as the main control system,Hikvision camera DS-2CD3T46FDWDV2-I8 S is used as the image acquisition system,and ABB's IRB6700 200/2.6 robot is used as the motion actuator.Under the unified coordination of the control software,we realize intelligent slag removal.2.For the identification and positioning of the slag image and the calculation of the slag content,the characteristic difference between the slag and the zinc liquid is analyzed,and the traditional image processing process is used,based on the Open CVSharp algorithm library,to study and realize the preprocessing and morphological analysis of the slag image.At the same time,the analysis results of the slag image by the edge detection operators of Prewitt,Sobel,Laplacian,and Canny are comparatively studied.Finally,the recognition and positioning of the slag is realized based on the Laplacian operator,and the calculation of the slag content in the area is realized by combining the analysis of variance.3.In view of the hand-eye calibration problem of the robot vision system of the galvanizing line,we analyze and determine the vision system structure to be Eye-to-Hand.By analyzing the camera's pinhole imaging mathematical model,we build an imaging model from space point to pixel coordinate point,and use Zhang's calibration method to achieve camera calibration,and obtain internal and external parameters and distortion parameters.By simplifying the zinc pot liquid level into a 2D scene,we use the ninepoint method to calculate the transformation matrix from the pixel coordinate system of the zinc pot pool image to the robot base coordinate system,and verify the accuracy of the transformation matrix to meet actual application requirements.4.Based on the above research work,the slag removal robot system based on machine vision was successfully applied to the galvanizing line.Based on the Win Form application form,we realize the development of the host computer control software by programming.Under the complex conditions of the galvanizing scene in a narrow space and a large span,the robot motion trajectory planning and one-key reset function are realized.Finally,the actual slag removal effect is analyzed and compared with the firstgeneration slag removal robot.The slag removal robot system based on machine vision improves the slag removal efficiency and achieves a half-hour slag removal rate of 95%. |
PDF Full Text Request