Issue << content << Simultaneous Prediction of Residual Stress and Hardness Based on the Barkhausen Noise Measurements


DOI:

2011. . No. 4(54) - Dec 2011, pp. 78-83

Simultaneous Prediction of Residual Stress and Hardness Based on the Barkhausen Noise Measurements

A. Sorsa, K. Leiviskä

Quantitative prediction of material properties based on the Barkhausen noise measurements is challenging. A data-based approach for building a prediction model consists of feature generation and  selection, model identification and validation of steps. In this study, a multitude of features is generated from the Barkhausen noise measurements taken from a case-hardened steel samples. The aim is to predict residual stress and hardness simultaneously based on these features. A simple forward-selection algorithm is used in feature selection. Features are selected for a multivariable linear regression model. The multivariable linear regression model with the selected features is identified. Throughout the selection and identification procedures k-fold cross-validation is used to guarantee that the results are realistic and hold also for future predictions. Prediction accuracy of the developed model is acceptable.

References
  1. Lindgren M., Lepistö T. Effect of prestraining on Barkhausen noise vs. stress relation. – NDT&E Internat. 2001. V. 34. P. 337 – 344.
  2. O'Sullivan D., Cotterell M., Tanner D. A.,Mészáros I. Characterisation of ferritic stainless steel by Barkhausen techniques. – NDT&E Internat. 2004. V. 37. P. 489 – 496.
  3. Stefanita C.-G., Atherton D. L., Clapham L. Plastic versus elastic deformation effects on magnetic Barkhausen noise in steel. – Acta Materialia. 2000. V. 48. No. 13. P. 3545 – 3551.
  4. Saquet O., Chicois J., Vincent A. Barkhausen noise from plain carbon steels: analysis of the infl uence of microstructure. – Materials Sci. & Eng. Ser. A. 1999. V. 269. P. 73 – 82.
  5. Blaow M., Evans J. T., Shaw B. A. Effect of hardness and composition gradients on Barkhausen emission in case hardened steel. – J. of Magnetism and Magnetic Materials. 2006. V. 303. P. 153 – 159.
  6. Stewart D. M., Stevens K. J., Kaiser A. B. Magnetic Barkhausen noise analysis of stress in steel. – Current Appl. Phys. 2004. V. 4. P. 308 – 311.
  7. Sorsa A., Leiviskä K., Santa-aho S. Prediction of residual stress from the Barkhausen noise signal. – In: NDT 2008 Conf. and Exhib., 15 – 18 Sep. – Cheshire, UK: BiNDT, 2008. – 10 p.
  8. Sorsa A., Leiviskä K. Feature Selection from Barkhausen Noise Data Using Genetic Algorithms with Cross-Validation» – In: Adaptive and Natural Computing Algorithms, 9th International Conference, ICANNGA 2009 / Ed. by M. Kolehmainen, P. Toivanen, B. Beliczynski. – Kuopio, Finland: April 23 – 25, 2009. Revised Selected Papers.– Berlin / Heidelberg: Springer, 2009, P. 213 – 222.
  9. Sorsa A., Leiviskä K., Santa-aho S. et al. Quality feature modeling of laser processed grinding burns based on Barkhausen noise measurements. – In: NDT09 Conf. and Materials Testing 2009 Exhibition. – Blackpool, UK: 2009, 15 – 17 Sept. – 10 p.
  10. Sorsa A, Leiviskä K. An entropy-based approach for the analysis of the Barkhausen noise signal. – In: Proc of 7th Internat. Conf. on Barkhausen Noise and Micromagnetic Testing. – Aachen, Germany, 2009, 15 – 16 July, 11 p.
  11. Kohavi R., John G. H. Wrappers for feature subset selection. – Artifi cial Intelligence. 1997. V. 97. P. 273 – 324.
  12. Baumann K. Cross-validation as the objective function for variable-selection techniques. – TrAC Trends in Analytical Chemistry. 2003. V. 22. P. 395 – 406.
  13. Guyon I., Elisseeff A. An introduction to variable and feature selection. – J. of Machine Learning Res. 2003. V. 3. P. 1157 – 1182.
  14. Wisnovski J. W., Simpson J. R., Montgomery D. C., Runger G. C. Resampling methods for variable selection in robust regression. – Computational Statistics & Data Analysis. 2003. V. 43. P. 341 – 355.
  15. Anderssen E., Dyrstad K., Westad F., Martens H. Reducing over-optimism in variable selection by cross-model validation. – Chemometrics and Intelligent Lab. Systems. 2006. V. 84. P. 69 – 74.
  16. Barros A. S., Rutledge D. N. Genetic algorithm applied to the selection of principal components. – Ibid. 1998. V. 40. P. 65 – 81.
  17. Depczynski U., Frost V. J., Molt K. Genetic algorithms applied to the selection of factors in principal component regression. – Analytica Chimica Acta. 2000. V. 420. P. 217 – 227.
  18. Benoudjit N., Cools E., Meurens M., Verleysen M. Chemometric calibration of infrared spectrometers: selection and validation of variables by non-linear models. – Ibid. 2004. V. 70. P. 47 – 53.
  19. Hemmateenejad B. Correlation ranking procedure for factor selection in PCANN modeling and application to ADMETox evaluation. – Ibid. 2005. V. 75. P. 231 – 245.
  20. Engelen S., Hubert M. Fast model selection for robust calibration methods. – Analytica Chimica Acta. 2005. V. 544. P. 219 – 228.
  21. Harrell F. E. Regression Modeling Strategies with Applications to Linear Models, Logistic Regression, and Survival Analysis. – New York: Springer, 2001.
  22. Sorsa A., Leiviskä K., Santa-aho S. et al. A study on laserprocessed grinding burn simulation and analysis on Barkhausen noise measurement. – Insight. 2010. V. 52. P. 293 – 297.
  23. Lei Y., He Z., Zi Y., Hu Q. Fault diagnosis of rotating machinery based on a new hybrid clustering algorithm. – Internat. J. of Adv. Manufacturing Techn. 2008. V. 35. P. 968 – 977.
  24. Smit S., Hoefsloot H. C. J., Smilde A. K. Statistical data processing in clinical proteomics. – J. of Chromatography B. 2008. V. 866. P. 77 – 88.
 
published

CONTACTS

5 Udelny Pr.
St. Petersburg
007 812 670 7609
editor@ndtworld.com


NDT World, Russian Quarterty Journal
©2000-2016 All rights reserved