• Optics and Precision Engineering
  • Vol. 32, Issue 18, 2783 (2024)
Guifang QIAO1,2,*, Chunhui GAO1, Xinyi JIANG1, Simin XU1, and Di LIU1
Author Affiliations
  • 1School of Automation, Nanjing Institute of Technology, Nanjing267, China
  • 2School of Instrument Science and Engineering, Southeast University, Nanjing10096, China
  • show less
    DOI: 10.37188/OPE.20243218.2783 Cite this Article
    Guifang QIAO, Chunhui GAO, Xinyi JIANG, Simin XU, Di LIU. Spatial error prediction method for industrial robot based on Support Vector Regression[J]. Optics and Precision Engineering, 2024, 32(18): 2783 Copy Citation Text show less
    Schematic diagram of the joint coordinate system for the Staubli TX60 industrial robot
    Fig. 1. Schematic diagram of the joint coordinate system for the Staubli TX60 industrial robot
    Industrial robot calibration system
    Fig. 2. Industrial robot calibration system
    Structure of SVR model
    Fig. 3. Structure of SVR model
    Spatial distribution of pose measurement points for Staubli TX60 industrial robot
    Fig. 4. Spatial distribution of pose measurement points for Staubli TX60 industrial robot
    Position error compensation for Staubli TX60 robot
    Fig. 5. Position error compensation for Staubli TX60 robot
    Attitude error compensation for Staubli TX60 robot
    Fig. 6. Attitude error compensation for Staubli TX60 robot
    Comparison of four methods for fitting position errors
    Fig. 7. Comparison of four methods for fitting position errors
    Comparison of four methods for fitting attitude errors
    Fig. 8. Comparison of four methods for fitting attitude errors
    [in Chinese]
    Fig. 9. [in Chinese]
    iθi/raddi/mmai/mmαi/radβi/rad
    1π00π/2-
    2π/2029000
    3π/2200π/2-
    4π3100π/2-
    5π00π/2-
    607000-
    Table 1. Theoretical MD-H parameters of Staubli TX60 robot
    Number of layerInputs LayerMiddle LayerOutputs Layer
    BP4620/206
    Elman4620/206
    Table 2. Neural network parameter settings
    ModelAverage errorMaximum errorStandard deviation of error
    Position/mmAttitude/(°)Position/mmAttitude/(°)Position/mmAttitude/(°)
    Origin0.706 10.174 21.194 70.280 40.276 00.051 0
    BP0.052 60.057 50.136 80.134 00.024 10.024 5
    Elman0.051 70.059 00.140 40.145 30.023 40.026 9
    LM0.196 80.065 90.587 60.201 80.083 20.037 5
    SVR0.055 60.024 60.152 30.107 30.023 90.017 5
    Table 3. Comparison experimental results of robot pose error compensation
    Guifang QIAO, Chunhui GAO, Xinyi JIANG, Simin XU, Di LIU. Spatial error prediction method for industrial robot based on Support Vector Regression[J]. Optics and Precision Engineering, 2024, 32(18): 2783
    Download Citation