Neural-fuzzy variable gap control method for GMAW pipe-line welding with CCD camera
He sensing of the weld pool and controlling of torch at the center of the groove are important problems in back welding of GMAW (Gas Metal Arc Welding) for pipeline, furthermore, the gap of the groove perhaps is varied, which needs an intelligent control strategy to obtain the high welding quality. Fuzzy neural network control method based on BP algorithm is proposed in this paper, from the module of image processing, the corresponding gap location and width can be obtained. Then determine corresponding swing width and speed when weld gap is varied by the network fuzzy inference and calculating Euclidean distance for GMAW variable gap backing welding process. Experiment results show that the designed control method can improve the welding quality compared with traditional fixed swing and the traditional auto swing.
Fuzzy neural network Variable gap BP algorithm Euclidean distance Seam tracking
Hu Hailin Li Jing Li Fang Zhu Wei Pang Heqiang
College of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics Nanji College of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics Nanji Industrial Robot Research Center, Industrial Technology Research Institute, Kunshan, Jiangsu Provinc
国际会议
合肥
英文
2358-2363
2011-09-23（万方平台首次上网日期，不代表论文的发表时间）