Research On Weld Seam Tracking System Based On Information Fusion Of Magnetron Arc-laser Sensor

Under the environment circumstance of Industry 4.0,welding automation and intelligence have become the trend of production in contemporary companies.The current increase in the size of welded components has become a general trend,At present,the large-scale welding component has become a general trend,which means that the application of narrow gap welding will become increasingly widespread.Taking into account the narrow gap welding process requirements at the groove,in order to improve the company’s production efficiency and the labor environment of workers,It is currently the development trend of narrow gap welding to improve the accuracy of tracking technology and optimize the structure of the welding torch.Therefore,the purpose of this study is to optimize the structure of the welding torch and enhance the tracking performance,and make the welding more intelligent.In this paper,narrow gap welding is taken as the research object.By conducting research about arc-laser visual inspection principle and information fusion technology,The collected magnetic arc and laser vision signals are fused by fuzzy theory,which makes the collected weld information more accurate,The fused information is transmitted to the MAG welding seam tracking module after welding.The welding system formulates the trajectory planning scheme according to the deviation information to control the welding gun’s path motion according to the trajectory planning,so the accuracy of welding seam and good side wall fusion are guaranteed at last,so that the information of the collected weld can be changed in real time,and then the information The weld tracking module passed to the MAG welding formulates the trajectory planning scheme based on the deviation information,controls the path movement of the welding torch according to the trajectory planning,and ultimately ensures the accuracy of the welding seam and good sidewall fusion.(1)In this paper,by studying magnetically controlled swing arc and three-line laser vision information acquisition technology,it is applied to magnetically controlled narrow gap MAG welding tracking,and the excitation magnetic field is used to control the left and right deflection directions of the MAG welding arc,and within a certain range.Affects the weld width range under the arc,providing feasibility for narrow gap welding under complex trajectories.(2)Studying the method of locating the visual feature points,adopting the three-wire laser to pass the mathematical relationship between the coordinate value of the two-dimensional pixel point and the three-dimensional mechanical coordinate value,and the function conversion can provide a lot of information to determine the weld spot.The interference caused by arc and splash during welding can be reduced,and the influence of arc and splash noise may cover important parts of laser stripe can be reduced,these impacts may lead to the failure of detection of weld feature points.In order to further improve the tracking accuracy,fuzzy inference information fusion methods are introduced.After the collected electrical parameters are converted by data,adaptive fuzzy control method is adopted to adjust the input and output fields,fuzzy rules and membership functions in real time,and the weld information is organically integrated to achieve automatic tracking welding.(3)By studying the weld information,the number of weld bead layers and the number of bead welds in each layer are estimated,and the trajectory planning theory is applied,from which we can obtain theoretical deviation values.In order to reduce the tracking error caused by lag factors,combined with the welding bead theoretical prediction program,according to the weld trajectory tangent fitting idea and the cross slide motion characteristics,the linear fitting idea is applied to obtain the weld seam.Motion prediction information,and the simulation obtained weld trajectory prediction to meet the tracking accuracy.(4)We verify the above-mentioned fusion trajectory prediction theory by constructing related experimental platforms.The welding performance of the weld shows that the weld has good predictability,stable work performance,good weld process,and high real-time performance.Forecasting requirements.