Tool Wear Estimation Using Support Vector Machines in Ball-nose End Milling

Sheng Huang, Xiang Li, and Oon Peen Gan
Submission Type: 
Full Paper
AttachmentSizeTimestamp
phmc_10_016.pdf271.81 KBSeptember 26, 2010 - 11:40pm

To reduce production cost and improve product quality, on-line tool wear monitoring are required from industry in the ball-nose end milling process. This paper will introduce a method to determine the tool wear by measured cutting force. The feature will be extracted from the measured cutting force with different flank wear. As the adaptive window width in wavelet transform is an advantage for analyzing and monitoring the rapid transient of small amplitude of cutting force signal when cutting engagement changes along the sculptured surface tool path, wavelet transform (WT) is more effective than FFT monitoring index for ball-nose end milling. In this research, cutting force signals will be analyzed in time-frequency domain to explore sensitive monitoring features in ball-nose end milling sculptured surfaces. Neural network approaches have been widely used in tool wear estimation because of their learning capability. As a supervised method, support vector machines (SVM) was developed for the classification problem to take advantage of prior knowledge of tool wear and construct a hyper-plane as the decision surface. In this paper, SVM will be formulated into regression problem to estimate tool wear rather than decision maker.

Publication Control Number: 
016
Submission Keywords: 
support vector machines
Tool condition monitoring
Submitted by: 
  
 
 
 

follow us

PHM Society on Facebook Follow PHM Society on Twitter PHM Society on LinkedIn PHM Society RSS News Feed