Hidden Markov Modelling of High-Speed Milling (HSM) Process Using Acoustic Emission (AE) Signature for Predicting Tool Conditions

Abstract

Tool condition monitoring is an important activity to monitor and maintain the quality of products manufactured in any machining process without any manual intervention. Hidden Markov models (HMM) are developed in this study for predicting tool conditions in a High-Speed Milling of titanium alloy using a carbide tool. Tool conditions are predicted using AE signatures captured during the metal cutting operation. A correlation between AE features and tool conditions were established using Baum-Welch and Viterbi algorithms. HMM models proposed in this study are integrated with the K-means clustering algorithm. The clustered data has been represented as an integer sequence and is divided into 3 tool states such as ‘sharp’, ‘intermediate’ and ‘worn-out’. Three HMM models are created for each state of the tool. Two AE features namely ‘Root Mean Square (RMS)’ and ‘Rise’ were used for developing HMMs. The performance of the HMMs is evaluated using log-likelihood measure.