A modular tool wear monitoring system in a metal cutting operation using MLP neural networks and multivariate process parameters

The application of multi-layer perceptron (MLP) neural networks to cutting tool wear classification in a metal turning operation is reported. Cutting tests were conducted using carbide inserts with and without wear on alloy steel, and the acquired multivariate data were used to train, validate and test the classification capabilities of two MLP configurations. Training was achieved via backpropagation of error enhanced by the addition of a momentum term and adaptive learning rate. Results of successful classification of the tool state ranged from 88-96%