• Optics and Precision Engineering
  • Vol. 29, Issue 11, 2649 (2021)
Xin-yuan WEI1, Yu-chen CHEN1, En-ming MIAO2,*, Xu-gang FENG1, and Qiao-sheng PAN3
Author Affiliations
  • 1School of Electrical and Information Engineering, Anhui University of Technology, Ma’anshan243032, China
  • 2School of Mechanical Engineering, Chongqing University of Technology, Chongqing400054, China
  • 3School of Instrument Science and Opto-electronics Engineering, Hefei University of Technology, Hefei20009, China
  • show less
    DOI: 10.37188/OPE.20212911.2649 Cite this Article
    Xin-yuan WEI, Yu-chen CHEN, En-ming MIAO, Xu-gang FENG, Qiao-sheng PAN. Application of principal component algorithm in spindle thermal error modeling of CNC machine tools[J]. Optics and Precision Engineering, 2021, 29(11): 2649 Copy Citation Text show less
    Experimental object: three axis vertical mechining center
    Fig. 1. Experimental object: three axis vertical mechining center
    Position of displacement sensor
    Fig. 2. Position of displacement sensor
    Temperature curve of K1
    Fig. 3. Temperature curve of K1
    Temperature curve of K9
    Fig. 4. Temperature curve of K9
    Thermal errors in Z direction curves of each experiment
    Fig. 5. Thermal errors in Z direction curves of each experiment
    The influence of the number of TSPs on the prediction effect of the model
    Fig. 6. The influence of the number of TSPs on the prediction effect of the model
    Structure of BP neural network
    Fig. 7. Structure of BP neural network
    Fitting accuracy results of four modeling algorithms
    Fig. 8. Fitting accuracy results of four modeling algorithms
    Prediction accuracy results of four modeling algorithms
    Fig. 9. Prediction accuracy results of four modeling algorithms
    Robustness results of four modeling algorithms
    Fig. 10. Robustness results of four modeling algorithms
    Verification experiment V1~V3 spindle speed setting
    Fig. 11. Verification experiment V1~V3 spindle speed setting
    The curve of V1 predicted by K1 model
    Fig. 12. The curve of V1 predicted by K1 model
    Thermal error compensation measurement results
    Fig. 13. Thermal error compensation measurement results
    传感器安放位置作用
    T1~T5主轴前轴承测量电机发热
    T7,T8主轴外箱测量主轴发热
    T6,T9主轴电机测量主轴发热
    T10机床外壳测量环境温度
    Table 1. Position and function of temperature sensor
    批次

    环境温度

    范围/℃

    主轴转速/r·min-1进给速度/mm·min-1
    K13.81-6.002 0001500
    K22.94-4.944 000
    K33.62-7.196 000
    K49.00-10.932 000
    K513.81-16.004 000
    K612.94-14.946 000
    K721.12-24.812 000
    K828.68-33.754 000
    K932.37-35.066 000
    Table 2. Experimental parameters of each batch
    ZiZ1Z2Z3Z4Z5Z6Z7Z8Z9Z10
    Vi98.51.350.110.020.0100000
    Table 3. Principal components and contribution rate of K1
    批次K1K2K3K4K5K6K7K8K9
    结果1,71,71,71,71,101,101,88,101,8
    Table 4. Selection results of TSPs of each experiment(PCA)
    批次K1K2K3K4K5K6K7K8K9
    结果1,101,101,101,81,101,101,81,101,10
    Table 5. Selection results of TSPs of each experiment(FCGC)
    批次模型系数
    b0b1b2
    K1-3.676.06-3.97
    K2-3.836.17-4.82
    ……
    K84.499.39-11.14
    K9-3.098.05-2.89
    Table 6. Model coefficient of MLR of each batch experiment
    批次模型系数
    b0b1b2
    K16.2581.8111.773
    K214.4741.6331.644
    ……
    K86.6562.2442.348
    K914.9051.7751.850
    Table 7. Model coefficient of ridge regression of each batch experiment
    Xin-yuan WEI, Yu-chen CHEN, En-ming MIAO, Xu-gang FENG, Qiao-sheng PAN. Application of principal component algorithm in spindle thermal error modeling of CNC machine tools[J]. Optics and Precision Engineering, 2021, 29(11): 2649
    Download Citation