The Research Of Magnetic-Control SAW Automatic Tracking System Based On The Variational Mode Decomposition

On the background of powerful manufacturing strategy of “Made in China 2025”, to achieve the automation and intelligence of welding has become rather imperative. The research and exploitation on advanced automatic welding equipment, as a result, has become an inevitable trend in welding area, while the key step is the extraction of signal deviation. In this paper, based on the characteristics of long straight weld in electric wheel dump truck compartment, a Magnetic-Control Submerged Arc Welding(SAW) tracking system was applied to achieve automatic welding. The Magnetic-Control SAW seam tracking system was analyzed in detail, the structure of Magnetic-Control SAW arc-sensor was improved as well. At the same time, the intelligent optimized algorithm was introduced to optimize parameter of Magnetic-Control SAW. Moreover, the tracking signal was analyzed by using variational mode decomposition method to improve the systematic constancy and tracking accuracy. The summary of results is listed below:1. Considered both characteristics of long straight weld in electric wheel dump truck compartment and welding conditions, the structure of Magnetic-Control Arc sensor was carefully re-designed. As a result, easier parameter adjustment and strengthened installation were achieved. Also the impact of the sensor itself on the tracking signal was reduced to a minimum scale. In conclude, the new structure of re-designed sensor effectively improved the stability of welding system.2. The intelligent optimized algorithm was applied to optimize parameter of the Magnetic-Control SAW due to the fact that the arc tracking signal is substantially influenced by variation of process parameters. The combination of genetic algorithmand and BP neural network used in this research optimized Magnetic-Control SAW parameters. Not only did this method cut down the number of experiments, but it also improved the efficiency and reliability in acquiring optimal process parameter group, which serves as a good foundation for the extraction of tracking signals.3. In order to analyze the filtered tracking signal containing varoius interfering signals, Variational Mode Decomposition(VMD) was introduced. Analysis results showed that Variational Modal Decomposition could accurately extract the effective tracking waveform from multi-interferential signals, which serves as a theoretical basis for the improvement of accurate seam tracking.4. To address the phase lag problem caused by filtering of the previous circuit, the previous filter circuit was carefully re-designed. Through combination of the second-order Butterworth low-pass filter circuit and phase-shift circuit, the phase lag phenomenon was significantly improved, as well as the accuracy and instantaneity of the tracking signal.5. An experiment was carried out on the optimized Magnetic-Control Arc tracking sensor system to compare the tracking effect with or without usage of Variational Mode Decomposition. The experiment indicated that VMD effectively improved the accuracy of seam tracking.