A NEURAL NETWORK APPROACH TO TOOL WEAR MONITORING IN END MILLING OPERATIONS

摘要

This paper outlines a methodology of continuously and indirectly monitoring tool wear in end milling operations. A neural network based computer control system was developed for monitoring the tool wear with machining variables such as spindle speed, feed rate, and depth of cut using measured cutting forces under a variety of cutting conditions in end milling operations. The supervised neural network developed was able to extract tool wear information from the changes occurring in the machining process. The results indicate that the neural networks approach presents an efficient and economic method for predicting the tool condition with the measured process variables.