The Research On The Key Technolgies Of Intelligent Monitoring System For GMAW Pipe-line Backing Welding Process

With the rapid development of the oil industry and pipeline, the requirements of high qualitysteel was still in the trend with the steady increase for high-pressure, high-intensity, long-distancetransmission. At the same time, since the development of the intelligence and automation of thewelding process, more and more researchers began to pay attention on the research of the weldingprocess quality control, and then the pipe-line welding process monitoring system is the key problemin welding quality control process to be solved.Firstly, the current development of GMAW (Gas metal arc welding) intelligent monitoringsystem based on multi-sensor for pipe-line welding is discussed. Secondly, gives the system design ofthis intelligent welding process monitor system and then the system architecture is presented. Thirdly,describes the multi-sensor information processing subsystem, which includes the extraction of weldseam deviation control parameter, the welding pool coarse positioning method, the sobeltransformation method to extract the welding seam position and the20-frame average method todetermine the final location of the upper and lower weld seam location; Fourthly, gets torch heightcorrection value and level reference value according to the linear relation between arc current andheight or level of weld torch; Fifthly, achieve visual image de-noising using the redundant informationfrom the CCD(Charg-coupled Device) sensor and the Arc sensor; Sixthly, describes the variable-gapseam tacking subsystem, after gets the weld gap based on CCD sensor, then a new control method foradjusting the welding torch combed with BP neural network and fuzzy control method is described;Lastly, achieving the arc torch height and level of bias control according to multi-sensor fusioninformation technology.Compared with the current similar research, the efficiency and accuracy of seam tacking isimproved. The results show that in the application of MIG(Metal Inert Gas Arc Welding) weldingwith much noise, the multi-sensor information processing sub-system can accurately extract the upperand lower weld seam location and the variable-gap welding torching correction subsystem can beadaptive in the control of the pipeline gap as well as achieve level with a high degree of correctionsimultaneously, and the welding results show that this system can obtain a better welding seam shape.