A neural network-based model for the prediction of cutting force in milling process. A progress study on a real case

In spite of recent developments focusing on milling process optimization through an effective cutting force control, there is a need for the analysis of the transient response of these systems because undesirable oscillations in cutting force can be harmful to the quality of the finishing surface and tools. The main goal of this work is to develop a versatile neural network model which can online predict the mean cutting force under commonly encountered conditions. Using this model, easily obtained from a straightforward machining test, developments of complex adaptive controllers and monitoring systems can be carried out. As a result, a good model for predicting the cutting process was obtained.