Prediction of tool wear using machine vision approach

Tool wear prediction plays a crucial role in the machining industry for proper planning and optimization of cutting conditions. Nevertheless, tool wear assessment method using sensor signals has its drawbacks in the industry application. The objective of this study is to apply Artificial Neura...

Full description

Bibliographic Details
Main Authors: Sim, Pei Chin, Lee, Woon Kiow, Abdullah, Haslina, Talib, Norfazillah, Ong, Pauline, Saleh, Aslinda, Ahmad, Said, Sung, Aun Naa
Format: Article
Language:English
Published: 2022
Subjects:
Online Access:http://eprints.uthm.edu.my/7148/
http://eprints.uthm.edu.my/7148/1/J14148_72aaaa163aaaa2172e5b3a0bf92821e7.pdf
Description
Summary:Tool wear prediction plays a crucial role in the machining industry for proper planning and optimization of cutting conditions. Nevertheless, tool wear assessment method using sensor signals has its drawbacks in the industry application. The objective of this study is to apply Artificial Neural Network (ANN) prediction model and machine vision system to predict flank wear in turning operation based on the texture images of machined surface captured by complementary metal oxide semiconductor (CMOS) camera in-cycle. The image pre-processing technique was utilized to enhance the quality of surface texture images acquired from the experiment and the texture descriptors were extracted from the processed images using gray-level co-occurrence matrix (GLCM). Three ANN prediction models with different input variables were developed using MATLAB software. The findings showed that the ANN prediction model with input variables of contrast, entropy, cutting speed, and feed rate outperformed the other ANN prediction model. The prediction accuracy of this model in estimating flank wear reached up to 93.18%. A very good fit and the relationship could be found in this model with R2 of 0.9863 for flank wear.