Tool Wear Monitoring System Using Artificial Neural Network

Murat Sönmez

Department of Electronics & Com. Engineering,
University of Kocaeli, Izmit,Turkey
e-mail: mrsonmez@kou.edu.tr

İsmet Kandilli

Industrial Electronics,
University of Kocaeli KMYO, Izmit, Turkey

H.Metin Ertunç

Assistant Professor, Department of Mechatronics Engineering,
University of Kocaeli, Izmit Turkey

Abstract

This paper outlines a neural network based modular tool condition monitoring system for cutting tool state classification. Tool wear estimation is based on the measurement of cutting force components. An important variation in mean, RMS, variance, standard deviation of cutting forces can result in estimation and classification error. Several methods to develop monitoring devices for observing the wear levels on the cutting tool on line while engaged in cutting have been attempted. In this work various sensors is used with a multilayer neural network for predicting the wear on the cutting tool. The neural network, trained off line using a backpropogation algorithm and the experimental data, is used to predict the wear.