PT Journal
AU Guo, S
   Zheng, H
   Liu, X
   Gu, L
TI Comparison on Milling Force Model Prediction of New Cold Saw Blade Milling Cutter Based on Deep Neural Network and Regression Analysis
SO Manufacturing Technology Journal
PY 2021
BP 456
EP 463
VL 21
IS 4
DI 10.21062/mft.2021.053
DE NCSBMC; MF; Orthogonal Test; Multiple Linear Regression Analysis; DNN
AB A four factors and three levels orthogonal milling force (MF) test is designed, which qualitatively obtains the influence of four factors, namely workpiece material, workpiece diameter, milling speed and feed per tooth, on MF of the new cold saw blade milling cutter (NCSBMC), then further verifies the reliability of test data with simulation analysis of MF. The multiple linear regression analysis and deep neural network (DNN) are used to accurately fit and predict the magnitude of MF in three directions of NCSBMC, taking into account the influence of workpiece material factors on MF. Compared with the results of empirical formula, DNN has higher prediction accuracy. The research results provide theoretical guidance for the optimization of milling parameters in actual machining process.
ER