Mean Shifts Monitoring of Auto-correlated Manufacturing Processes Using Long Short-Term Memory-Based Recurrent Neural Network

  • Shu-mei Chen
  • Jian-bo YuEmail author
Conference paper


The traditional statistical process control (SPC) technique is used as a monitoring tool to recognize mean shifts of discrete manufacturing processes. The fundamental assumption using SPC is the independence of observed process data. Such assumption is usually violated in practical industries, which proceeds the development of alternative schemes to monitor auto-correlated processes. This paper approaches a long short-term memory (LSTM)-based model to recognize mean shifts of auto-correlated processes with different autocorrelation coefficients and shift parameters. The performance of the proposed method is evaluated by an average run length of time series residual control chart in comparison with other monitoring schemes. It was observed that the LSTM-based monitoring approach could improve significantly upon the early detection of out-of-control operating cases in auto-correlated processes.


Auto-correlation Long short-term memory LSTM Statistical process control Shifts recognition 



This research was supported by the National Natural Science Foundation of China (No. 51375290, 71777173), the Fundamental Research Funds for Central Universities and Shanghai Science.


  1. 1.
    C.W. Zobel, D.F. Cook, Q.J. Nottingham, An augmented neural network classification approach to detecting mean shifts in correlated manufacturing process parameters. Int. J. Prod. Res. 42(4), 741–758 (2004)CrossRefGoogle Scholar
  2. 2.
    D. Montgomery, D.C. Friedman, Statistical Process Control in Computer Integrated Manufacturing (Dekker, New York, 1989)Google Scholar
  3. 3.
    D.F. Cook, C.-C. Chiu, Using radial basis function neural networks to recognize shifts in correlated manufacturing process parameters. IIE Trans. 30(3), 227–234 (1998)Google Scholar
  4. 4.
    D.F. Cook, C.W. Zobel, Q.J. Nottingham, Utilization of neural networks for the recognition of variance shifts in correlated manufacturing process parameters. Int. J. Prod. Res. 39(17), 3881–3887 (2001)CrossRefGoogle Scholar
  5. 5.
    I. Sutskever, O. Vinyals, Q.V. Le, Sequence to sequence learning with neural networks. Adv. Neural Inf. Process. Syst. 3104–3112 (2014)Google Scholar
  6. 6.
    G.E. Dahl, D. Yu, L. Deng, A. Acero, Context-dependent pre-trained deep neural networks for large-vocabulary speech recognition. IEEE Trans. Audio Speech Lang. Process. 20(1), 30–42 (2012)CrossRefGoogle Scholar
  7. 7.
    D. Ciresan, U. Meier, J. Schmidhuber, Multi-column deep neural networks for image classification, in CVPR (2012), pp. 3642–3649Google Scholar
  8. 8.
    H. Sak, A. Senior, F. Beaufays, Long short-term memory based recurrent neural network architectures for large vocabulary speech recognition, no. Cd (2014)Google Scholar
  9. 9.
    F.A. Gers, J. Schmidhuber, LSTM recurrent networks learn simple context free and context sensitive languages. IEEE Trans. Neural Netw. 12(6), 1333–1340 (2001)CrossRefGoogle Scholar
  10. 10.
    H. Palangi et al., Deep sentence embedding using long short-term memory networks: analysis and application to information retrieval. IEEE/ACM Trans. Audio Speech Lang. Process. 24(4), 694–707 (2016)CrossRefGoogle Scholar
  11. 11.
    H. Sepp, J. Schmidhuber, Long short-term memory. Neural Comput. (1997)Google Scholar
  12. 12.
    A. Graves, J. Schmidhuber, Framewise phoneme classification with bidirectional LSTM networks, in Proceedings of International Joint Conference on Neural Networks, vol. 4 (2005), pp. 2047–2052Google Scholar
  13. 13.
    T.P. Longnecker, M.T. Ryan, Charting correlated process data (1993)Google Scholar
  14. 14.
    J. Yu, J. Liu, LRProb control chart based on logistic regression for monitoring mean shifts of auto-correlated manufacturing processes. Int. J. Prod. Res. 49(8), 2301–2326 (2011)CrossRefGoogle Scholar
  15. 15.
    D.G. Wardell, H. Moskowitz, R.D. Plante, Run-length distributions of special-cause control charts for correlated processes. Technometrics 36(1), 3–17 (1994)CrossRefGoogle Scholar
  16. 16.
    H.B. Hwarng, Detecting process mean shift in the presence of autocorrelation: a neural-network based monitoring scheme. Int. J. Prod. Res. 42(3), 573–595 (2004)CrossRefGoogle Scholar

Copyright information

© Springer Nature Singapore Pte Ltd. 2019

Authors and Affiliations

  1. 1.School of Mechanical and Energy EngineeringUniversity of TongjiShanghaiChina

Personalised recommendations