Title: Application of Support Vector Regression in assessment of geometric errors

Authors: M.S. Shunmugam, Vidhya Venkat

Addresses: Department of Mechanical Engineering, Indian Institute of Technology Madras, Chennai-600036, India. ' Department of Mechanical Engineering, Indian Institute of Technology Madras, Chennai-600036, India

Abstract: Manufactured components are verified using Coordinate Measuring Machine (CMM) and form-measuring instruments or setups. Standards define form error based on a Minimum Zone (MZ) principle which establishes an ideal reference feature. This principle is extended in the present work to arrive at the relationship errors. Additionally, idealisation of relationship has been attempted for the first time. Support Vector Regression (SVR) technique is used to evaluate form and relationship errors from CMM and form data. For new data sets and data taken from literature, the results show that SVR method can produce unique parameters for MZ.

Keywords: minimum zone; support vector regression; SVR; form error; geometric relationship; form data; CMM data; coordinate measuring machines; CMMs; geometric errors.

DOI: 10.1504/IJMR.2008.017413

International Journal of Manufacturing Research, 2008 Vol.3 No.2, pp.149 - 171

Published online: 06 Mar 2008 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article