Chinese Journal of Lasers, Volume. 49, Issue 16, 1602011(2022)

Online Weld Width Detection of Laser-MIG Hybrid Welding Based on Kalman Filter Algorithm Compensated by BP Neural Network

Xiuhang Liu, Yuhui Huang, Yanxi Zhang, and Xiangdong Gao*
Author Affiliations
  • Guangdong Provincial Welding Engineering Technology Research Center, Guangdong University of Technology, Guangzhou 510006, Guangdong, China
  • show less
    Figures & Tables(9)
    Laser-MIG hybrid welding process detection system
    Weld forming width extraction with second-order difference method
    Laser-MIG hybrid welding process image captured by high-speed camera
    Three-dimensional grayscale image of welding process
    Processing of laser -MIG hybrid welding molten pool image. (a) Extraction of region of interest (ROI); (b) mark of a molten pool; (c) image segmentation with watershed algorithm; (d) morphological processing
    Comparison of measured and true weld widths
    Comparison of BP neural network compensation Kalman filter (BP-KF) and Kalman filter (KF) values
    Absolute error values between true weld widths and the estimated value of BP-KF and KF
    • Table 1. Error statistical analysis

      View table

      Table 1. Error statistical analysis

      ErrorMAX /mmMAE /mmRMSE /mmMAPE /%
      Measurement error2.130.330.477.1
      KF error0.670.150.203.2
      BP-KF error0.650.130.162.7
    Tools

    Get Citation

    Copy Citation Text

    Xiuhang Liu, Yuhui Huang, Yanxi Zhang, Xiangdong Gao. Online Weld Width Detection of Laser-MIG Hybrid Welding Based on Kalman Filter Algorithm Compensated by BP Neural Network[J]. Chinese Journal of Lasers, 2022, 49(16): 1602011

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category: laser manufacturing

    Received: Oct. 19, 2021

    Accepted: Dec. 22, 2021

    Published Online: Jul. 28, 2022

    The Author Email: Gao Xiangdong (gaoxd666@126.com)

    DOI:10.3788/CJL202249.1602011

    Topics