Effect Of Welding Parameters Of CO₂ On The Change Of The Area Near The Weld Pool

Welding played the role of connecting workpieces in industrial production and processing,Therefore,welding is an indispensable part in processing and manufacturing.The reliability and stability of the performance of welding products was looked forward higher and higher with the rapid development of science and technology,Intelligent welding can improve the working environment of workers,the processing efficiency,and quality accuracy was getting higher and higher in the welding process,so intelligent welding was an inevitable trend of future development.Aimed at the metal in vicinity of weld pool splashing violently in CO₂ welding droplets short-circuit transition process,which affected the real-time and reliability of weld feature extraction seriously.In this paper,short-circuit transition CO₂ welding was taken as the object to study the variation law and characteristics of light intensity in vicinity of weld pool during carbon dioxide welding.Firstly,the relationship between image illumination and welding parameters in the vicinity of CO₂ melting pool was studied.The mapping relations between the welding parameters and the grayscale value in vicinity of weld pool was established with orthogonal test and improved BP neural network.The accuracy of the model has been analyzed and turned out to meet the accuracy requirements.The experimental results show that the experimental value was very close to that predicted by BP neural network.The BP neural network model can well reflect the relationship between the grayscale value of the image in the vicinity of the weld pool and the welding parameters.Secondly,The relationship between the grayscale value of the image in vicinity of weld pool and welding transient current was studied.Comparing the advantages and disadvantages of Lagrange interpolation method and least squares curve fitting method.When using the Lagrangian interpolation method,the number of interpolated polynomials was too high and both sides of the fitting curve would shock and Runge phenomenon occurred.Therefore,the mathematic model of average grayscale value of the image in the vicinity of the weld pool and welding transient current was established with the least squares method,The mathematical model satisfied to meet the accuracy requirement and revealed the law of arc changing during welding.Finally,the relationship between the grayscale value of the image in vicinity of weld pool and distance was studied.The data of points in different distance were collected with the system sampling method,image grayscale value and distance were fitted with cftool toolbox in Matlab,the mathematic model of grayscale value of the imagel and the change of distance was established,and the variation law of light intensity at different locations during welding was revealed.The influence of CO₂ welding parameters on the light intensity in the vicinity of the weld pool was studied,which lays the foundation of intelligent development of welding.