Advertisement

Diagnosis of Out-of-Control Signals in Complex Manufacturing Processes

  • Marcin PerzykEmail author
  • Jacek Kozlowski
  • Agnieszka Rodziewicz
Chapter
Part of the Studies in Systems, Decision and Control book series (SSDC, volume 183)

Abstract

Some new perspectives of applications of advanced data-driven modeling in control and fault diagnosis of manufacturing processes are presented. The time-series analysis can help to identify and isolate autocorrelations in the process, being a source of misleading conclusions about the process disturbances. Learning systems such artificial neural networks and classification trees can be used in identification of non-standard out-of-control signals. Finding the root-cases of the process disturbances can be facilitated using advanced models linking the process inputs and process outputs.

Notes

Acknowledgements

The authors would like to thank prof. Jan Jezierski and prof. Jan Szajnar for the permission to use the copyrighted material from our three papers which appeared in Archives of Foundry Engineering.

We would also like to thank prof. Witold Bialy for the permission to use the copyrighted material from our chapter in the monograph “Systems Supporting Production Engineering”, published by PA NOVA SA. Gliwice, Poland in 2013 (ISBN 978-83-937845-0-9).

References

  1. 1.
    Stanley, G.M.: Guide to Fault Detection and Diagnosis. White Paper available from: http://gregstanleyandassociates.com/whitepapers/FaultDiagnosis/faultdiagnosis.htm (2013)
  2. 2.
    Montgomery, D.C., Keats, J.B., Runger, G.C., et al.: Integrating statistical process control and engineering process control. J. Qual. Technol. 26, 79–87 (1994)CrossRefGoogle Scholar
  3. 3.
    Jiang, W., Farr, J.V.: Integrating SPC and EPC methods for quality improvement. Qual. Technol. Quant. Manag. 4/2007, 345–363, StatSoft Inc. (2014)Google Scholar
  4. 4.
    Statistica (data analysis software system) v.12. www.statsoft.com. Accessed 13 Apr 2016
  5. 5.
    Hoyer, R.W., Ellis, W.C.: A Graphical Exploration of SPC Part 2: the probability structure of rules for interpreting control charts. Qual. Prog. 29(5), 57–64 (1996)Google Scholar
  6. 6.
    Alwan, L.C., Roberts, H.V.: Time-series modelling for statistical process control. J. Bus. Econ. Stat. 6(1), 87–95 (1988)Google Scholar
  7. 7.
    Price, B., Price, K., et al.: SPC modifications for continuous autocorrelated processes. Manuf. Rev. 5, 184–192 (1992)Google Scholar
  8. 8.
    Berthouex, P.M., Hunter, W.G., et al.: Monitoring sewage-treatment plants—some quality control aspects. J. Qual. Technol. 10, 139–149 (1978)CrossRefGoogle Scholar
  9. 9.
    Harris, T.J., Ross, W.H.: Statistical process control procedures for correlated observations. Can. J. Chem. Eng. (1991).  https://doi.org/10.1002/cjce.5450690106CrossRefGoogle Scholar
  10. 10.
    Montgomery, D.C., Mastrangelo, C.M.: Some statistical process control methods for autocorrelated data. J. Qual. Technol. 23, 179–193 (1991)CrossRefGoogle Scholar
  11. 11.
    MacGregor, J.F., Harris, T.J.: The exponentially weighted moving variance. J. Qual. Technol. 25, 106–118 (1993)CrossRefGoogle Scholar
  12. 12.
    Wardell, D.G., Moskowitz, H., et al.: Control charts in the presence of data correlation. Manag. Sci. 38, 1084–1105 (1992)CrossRefGoogle Scholar
  13. 13.
    Chiu, C.C., Shao, Y.E., et al.: Identification of process disturbance using SPC/EPC and neural networks. J. Intell. Manuf. 14, 379–388 (2003)Google Scholar
  14. 14.
    Shao, Y.E., Wu, C.H., et al.: Identifying the change point of a process with the integration of SPC charts and neural networks. In: Proceedings of the Second International Conference on Innovative Computing, Information and Control ICICIC ‘07 Kumamoto Japan, pp. 400–403 (2007)Google Scholar
  15. 15.
    Zobel, C.W., Cook, D.F., et al.: An augmented neural network classification approach to detecting mean shifts in correlated manufacturing process parameters. Int. J. Prod. Res. 42, 741–758 (2004)CrossRefGoogle Scholar
  16. 16.
    Perzyk, M., Kozlowski, J., et al.: Application of computational intelligence methods in control and diagnosis of production processes. In: Kazmierczak, J. (ed.), Systems Supporting Production Engineering Gliwice Poland: PA NOVA SA, pp. 104–125 (2013)Google Scholar
  17. 17.
    Masters, T.: Practical Neural Network Recipes in C++, pp. 248–249. Academic Press (1993)Google Scholar
  18. 18.
    Perzyk, M., Krawiec, K., et al.: Application of time-series analysis in foundry production. Arch. Foundry Eng. 9(3), 109–114 (2009)Google Scholar
  19. 19.
    Perzyk, M., Maciejak, S., et al.: Application of time-series analysis for prediction of molding sand properties in production cycle. Arch. Foundry Eng. 11(2), 95–100 (2011)Google Scholar
  20. 20.
    Perzyk, M., Rodziewicz, A.: Application of time-series analysis in control of chemical composition of grey cast iron. Arch. Foundry Eng. 12(4), 171–175 (2012)Google Scholar
  21. 21.
    Green, M.E.: Critical Assessment of Current Metalcasting Green Sand System Control and Monitoring Processes (A Thesis in Industrial Engineering). The Pennsylvania State University, The Graduate School Harold and Inge Marcus Department of Industrial and Manufacturing Engineering (2009)Google Scholar
  22. 22.
    Hwarng, H.B., Hubele, N.F.: Back propagation pattern recognizers for X control charts: methodology and performance. Comput. Ind. Eng. 24, 219–235 (1993)CrossRefGoogle Scholar
  23. 23.
    Jacobs, D.A., Luke, S.R.: Training artificial neural networks for statistical process control. In: Proceedings of the Tenth Biennial University/Government/Industry Microelectronics Symposium, pp. 235–239.  https://doi.org/10.1109/ugim.1993.297059
  24. 24.
    Perzyk, M., Biernacki, R., et al.: Data mining in manufacturing: significance analysis of process parameters. J. Eng. Manuf. 222, 1503–1516 (2008)CrossRefGoogle Scholar
  25. 25.
    Perzyk, M., Kochanski, A.: Detection of causes of casting defects assisted by artificial neural networks. J. Eng. Manuf. 217, 1279–1284 (2003)CrossRefGoogle Scholar
  26. 26.
    Zhang, G.: A new diagnosis theory with two kinds of quality. Total Qual. Manag. 1(2), 249–257 (1990)CrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  • Marcin Perzyk
    • 1
    Email author
  • Jacek Kozlowski
    • 1
  • Agnieszka Rodziewicz
    • 1
  1. 1.Faculty of Production EngineeringWarsaw University of TechnologyWarsawPoland

Personalised recommendations