Experimental Simulation Of Automatic Slag Removing For Ladle Desulphurization Based On 3D Machine Vision

With the increase of steel production,the removing demand for desulphurization slags of iron ladle is getting higher and higher.Removing the slag of iron ladle is to eliminate the high sulfur residue after desulfurization treatment,the operation determines the quality of molten steel,thereby affecting the quality of the final steel product.At present,the desulphurization ladle is used to remove slag by manual operation,so the efficiency of slag scraping is low,moreover,the site environment is harsh and full of dust,so it is not conducive to the health and safety of operators.Based on the three-dimensional reconstruction technology,this thesis studies the simulation of the automatic remove slag of iron ladle after desulfurization.The research content of this thesis is as follows:(1)Design of automatic remove slag control system for desulphurization of iron ladle based on machine vision.It mainly includes two parts of the three-dimensional reconstruction design of the surface of the ladle and the automatic control design of the manipulator.According to the accuracy requirements of the three-dimensional reconstruction of the ladle,the movement space of the manipulator,and the space range of the ladle,the key system parameters of the automatic slag control system are determined.(2)Surface space reconstruction of ladle slag based on structured light.The premise of realizing the surface reconstruction of ladle steel slag is the accurate calibration of the camera and structured light.the internal and external parameters of the camera are calibrated,the factors affecting the accuracy of the camera calibration are analyzed,and the solution is proposed and verified;The line structure optical plane calibration algorithm is based on the checkerboard,the experimental results show that the algorithm has high precision and good performance;the practicality of the sub-pixel extraction of the structured light strip through the Hessian matrix,to obtain accurate structural light signals,to provide effective depth information for steel slag positioning;to obtain the spatial position of the steel slag surface using the least squares method,and verify the accuracy of the reconstruction plane.(3)Steel slag identification and positioning and automatic remove slag.The steel slag is identified according to the color and shape feature of the simulated steel slag,and the spatial position of the steel slag is obtained according to the three-dimensional reconstruction.Using the singular value decomposition to calibrate the camera coordinate system-manipulator coordinate system of the automatic slag system.(4)Experimental simulation of automatic remove iron slag after desulfurization.The steel slag surface is simulated with a plane placed randomly in space,and the foam blocks placed in the plane simulate steel slag,the experimental simulation of automatic slag stripping in iron ladle is carried out.Based on the automatic remove slag control system,the in-plane simulated steel slag is identified by image processing,and the three-dimensional spatial positioning detection of the simulated steel slag is carried out.For the steel slag planning slag space path,according to slag path,manipulator is moving,and finally the automatic slag is controlled,in the end,the effect is analyzed.Based on the active light structure light,the surface space reconstruction of the ladle is carried out.The automatic slag slag simulation of desulfurization of the iron ladle is carried out with the dobot manipulator.The final slag remove rate is 80%～90%.