Research on temperature field of milling insert with 3D complex groove

Prediction of temperature field is a key technology to achieve the groove design and reconstruction of milling insert, predictive model of neural network is a new way to achieve the prediction of temperature field. According to the non-steady state characteristic of temperature field of milling insert, the paper puts forward a predictive model of temperature field of milling insert with 3D complex groove based on Levenberg-Marquardt algorithm of BP neural network, and it overcomes the disadvantage that traditional neural network is easy to fall into local minimum. The predictive results show that this predictive model can converge quickly and predict accurately.