Skip to main content
SpringerLink
Log in

Anomaly detection for process monitoring based on machine learning technique

  • Original Article
  • Published:
Neural Computing and Applications Aims and scope Submit manuscript

Abstract

Anomaly detection is critical to process modeling, monitoring, and control since successful execution of these engineering tasks depends on access to validated data. The industrial process is uncertain in several situations, and the available information is formalized in terms of intervals. This article deals with the diagnostic of uncertain systems by multivariate static analysis. Linear Principal Component Analysis (PCA) and nonlinear Kernel PCA (KPCA) are generally used to deal with certain systems; they exploit single-valued variables. While in real situations these data are marred by uncertainties, these uncertainties cause difficulties in making decision in relation to the presence of defects. Thus, we have studied a recent and robust solution which consists in capturing the variability of multivariate observations by interval variables. In the first part, we treated a fault detection strategy based on interval PCA in the case of static linear systems. It includes first of all a comparative study between the deferent methods of detection of faults with interval PCA in which we proposed a new detection statistics of faults. In the second part, we studied a fault detection strategy based on interval KPCA method; we propose a reduction approach to solve the problem of nonlinearity and uncertainty and the problem of large data. The proposed fault detection methods are illustrated by synthetic data with an in-depth study and comparison using simulations of the air quality monitoring network and the Tennessee Eastman process.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9
Fig. 10
Fig. 11
Fig. 12
Fig. 13
Fig. 14
Fig. 15
Fig. 16
Fig. 17
Fig. 18
Fig. 19
Fig. 20
Fig. 21
Fig. 22
Fig. 23

Similar content being viewed by others

Abbreviations

VAR:

The variance

COV:

The covariance

Moy:

The average

M :

Number of system variables

\(N\) :

Number of samples

\(\ell\) :

Number of selected

R :

Number of reduced of samples

\(\left[ X \right]\) :

Interval data matrix

\(X^{{\text{m}}}\) :

Center matrix

\(X^{{\text{r}}}\) :

Range matrix

\(X_{{\text{R}}}\) :

The reduced data matrix

\({\text{LB}}\) :

Lower bound

\({\text{UB}}\) :

Upper bound

\(\overline{K}^{{\text{R}}}\) :

Reduced kernel matrix for UB

\(\underline{K}^{{\text{R}}}\) :

Reduced kernel matrix for LB

\(K_{{{\text{Rmr}}}}\) :

Reduced kernel matrix based on center and range

References

  1. Bounoua W, et al (2019) Online monitoring scheme. Using principal component analysis through Kullback-Leibler. Divergence analysis. Technique for fault detection. Trans Inst Meas Control 57–101.

  2. Russell EL, Chiang LH, Braatz RD (2012) Data-driven methods for fault detection and diagnosis in chemical processes. Springer, New York

    Google Scholar 

  3. Pearson K (1901) On lines and planes of closest fit to systems of points in space. Lond Edinb Dublin Phylosophical Mag J Sci 6:559–572

    MATH  Google Scholar 

  4. Hotelling H (1947) Techniques of statis-tical analysis- multivariate quality control-illustrated by air testing of sample bombsights. Mcgraw-Hill, New York, pp 11–148

    Google Scholar 

  5. Jolliffe IT (2002) Principal component analysis. Springer series in statistics. Springer, New York

    Google Scholar 

  6. Jackson JE (1991) A users guide to principal components and sons. Wiley, New Jersey

    Book  Google Scholar 

  7. Rao CR (1964) The use and interpretation of principal component analysis in applied research. Sankhyā Indian J Stat 26:329–358

    MATH  Google Scholar 

  8. Harkat MF, Mourot G, Ragot J (2006) An improved pca scheme for sensor fdi: Application to an air quality monitoring network. J Process Control 16:625–634

    Article  Google Scholar 

  9. Ku W, Storer R, Storer H, Georgakis C (1995) Disturbance detection and isolation by dynamic principal component analysis. Chemom Intell Lab 30:179

    Article  Google Scholar 

  10. Qin SJ (2003) Statistical process monitoring: basics and beyond. J Chemom 17:480–502

    Article  Google Scholar 

  11. Tulsyan A, Barton PI (2017) Interval enclosures for reachable sets of chemical kinetic flow systems. Part 1: Sparse transformation. Chem Eng Sci 166:334–344

    Article  Google Scholar 

  12. D’Urso P, Giordani P (2004) A least squares approach to principal component analysis for interval valued data. Chemometr Itell Lab Syst 70(179):192

    Google Scholar 

  13. Gioia P, Lauro C (2006) Principal component analysis on interval data. Comput Satat 21:343–363

    Article  MATH  Google Scholar 

  14. Irpino A (2006) “Spaghetti” PCA analysis: an extension of principal components analysis to time dependent interval data. Pattern Recognit Lett. 27:504–513

    Article  Google Scholar 

  15. Cazes P et al (1997) Extension de l’analyse en composantes principales à des données de type intervalle. Stat Appl 45(3):5–24

    Google Scholar 

  16. Chouakria A (1998) Extension des méthodes d'analyse factorielle à des données de type intervalle. Ph.D. dissertation, Université Paris-Dauphine, vol 6. pp 414,415,424,425

  17. Lauro CN, Palumbo F (2000) Principal component analysis of interval data: a symbolic data analysis approach. Comput Stat 15(1):73–78

    Article  MATH  Google Scholar 

  18. Le-Rademacher J, Billard L (2012) Symbolic covariance principal component analysis and visualization for interval-valued data. J Comput Graph Stat 21(2):413–432

    Article  Google Scholar 

  19. Ait-Izem T, et al (2017a) Approche neuronale d’ACP par intervalle appliquèe au diagnosti. In: (Quali 12 ème coninternational pluridisciplinaire en qualité, sûreté de fonctionnement et développement durable, Bourges- France

  20. Ait-Izem T et al (2017) Sensor fault detection based on principal component analysis for interval-valued data. Qual Eng 11:1–13

    Google Scholar 

  21. Plumbo F, Lauro NC (2003) A PCA for interval-valued data based on midpoints and radii. In New developments in psychometrics. Springer, Tokyo

  22. Ait-Izem T et al (2018) On the application of interval pca to process monitoring: a robust strategy for sensor fdi with new efficient control statistics. J Process Control 13:29–46

    Article  Google Scholar 

  23. Jaffel I et al (2016) Moving window KPCA with reduced complexity for nonlinear dynamic process monitoring. ISA Trans 64:184–192

    Article  Google Scholar 

  24. Taouali O et al (2015) New fault detection method based In reduced kernel principal component analysis(RKPCA). Int J Adv Manuf Technol 15:1547–1552

    Google Scholar 

  25. Harakat MF (2003) Détection et localisation de défauts par analyse en composantes principales. Thèse de doctorat de l’Institut National Polytechnique de Lorraine

  26. Harakat MF (2003) Détection et localisation de défauts par analyse en composantes principales. Thèse de doctorat l’Institut National Polytechnique de Lorraine

  27. Costa AQ, Pimentel B, Souza R (2010) K-means clustering for symbolic interval data based on aggregated kernel functions, tools with artificial intelligence (ICTAI). In: 22nd IEEE international conference IEEE. pp 375–379

  28. Costa A, Pimentel B, Souza R (2013) Clustering interval data through kernel-induced feature space. J Intell Inf Syst 40:109–140

    Article  Google Scholar 

  29. Pimentel B, Costa A, Souza R (2011) A partitioning method for symbolic interval data based on kernelized metric. In: Proceedings of the 20th ACM. International conference on Information and knowledg management, ACM. pp 2189–2191

  30. Hamrouni I et al (2020) Fault detection of uncertain nonlinear process using reduced interval kernel principal component analysis (RIKPCA). Int J Adv Manuf Technol

  31. Lahdhiri H et al (2017) Nonlinear process monitoring based on new reduced Rank-KPCA method. Stoch Environ Res Risk Assess 16:1833–1848

    Google Scholar 

  32. Jaffel I, Taouali O, Harkat MF (2016) Fault detection and isolation in nonlinear. Systems with partial reduced kernel principal component analysis method. Trans Inst Meas Control 40:1289–1296

    Article  Google Scholar 

  33. Chakour C, Benyounes A, Boudiaf M (2018) Diagnosis of uncertain nonlinear systems using interval kernel principal components analysis: application to a weather station. ISA Trans 83:126–141

    Article  Google Scholar 

  34. Harkat MF et al (2019) Fault detection of uncertain nonlinear process using interval-valued data-driven approach. Chem Eng Sci 205:36–45

    Article  Google Scholar 

  35. Mansouri M et al (2020) Data-driven and model-based methods for fault detection and diagnosis [Rapport]. Elsevier, New York

    Google Scholar 

  36. Wang H, Guan R, Wu J (2012) CIPCA: complete-information-based principal component analysis for interval-valued data. Neurocomputing 86:158–169

    Article  Google Scholar 

  37. Ait Izem T, et al (2015) Vertices and centers principal component analysis for fault detection and isolation. In: 2nd International conference on automationcontrol, engineering and computer science, Sousse-Tunisia

  38. Box G (1954) Some theorems on quadratic forms applied in the study of analysis of variance problems, I. Effect of inequality of variance in the one-way classification. Ann Math Stat 20:290–302

    Article  MATH  Google Scholar 

  39. Carlos F, Alaca S, Joe Q (2010) Reconstruction-based contribution for monitoring with kernel principal component analysis. Trans Inst Meas Control 17:7849–7857

    Google Scholar 

  40. Yanjie L, et al (2020) The instrument fault dection and identification based on Kernel principal component analysis and coupling analysis in process industry. Trans Inst Meas Control

  41. Scholkopf B et al (1998) Kernel pca pattern reconstruction via approximate pre-image. ICANN 98:147–152

    Article  Google Scholar 

  42. Aizerman M, Braverman E, Rozonoer L (1964) Theoretical foundations of the potential function method in pattern recognition learning. Autom Remote Control 98:821–837

    MATH  Google Scholar 

  43. Choi SW et al (2005) Fault detection and identification of nonlinear processes based on kernel PCA. Chemom Intell Lab Syst 75:55–67

    Article  Google Scholar 

  44. Harkat MF (2018) Fault detection of uncertain nonlinear process using interval-valued data-driven approach. Chem Eng Sci 14

  45. Nomikos P, MacGregor JF (1995) Multivariate SPC charts for monitoring batch processes. Technometrics 37:41–59

    Article  MATH  Google Scholar 

  46. Alcala CF, Qin SJ (2010) Reconstruction based conntribution for process monitoring with kernel principal component analysis. Ind Eng Chem Res 19:7849–7857

    Article  Google Scholar 

  47. Cui P, Li J, Wang G (2008) Improved kernel principal component analysis for fault detection. Expert Syst Appl 23:1210–1219

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Electronics and Micro-Electronic Laboratory, Faculty of Sciences of Monastir, National Engineering School of Sousse, University of Sousse, Sousse, Tunisia

    Imen Hamrouni

  2. National Engineering School of Monastir, University of Monastir, Monastir, Tunisia

    Hajer Lahdhiri & Okba Taouali

  3. Faculty of Computers and Information Technology, University of Tabuk, Tabuk, Saudi Arabia

    Khaoula Ben Abdellafou, Ahamed Aljuhani & Okba Taouali

Authors
  1. Imen Hamrouni
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Hajer Lahdhiri
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Khaoula Ben Abdellafou
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Ahamed Aljuhani
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Okba Taouali
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Okba Taouali.

Ethics declarations

Conflict of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendix

Appendix

See Table 8

Table 8 Cost of KPCA, RRKPCA, IKPCA_UL, IKPCA_CR, RRIKPCA_CR, RRIKPCA_UL
Full size table

.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Hamrouni, I., Lahdhiri, H., Ben Abdellafou, K. et al. Anomaly detection for process monitoring based on machine learning technique. Neural Comput & Applic 35, 4073–4097 (2023). https://doi.org/10.1007/s00521-022-07901-2

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00521-022-07901-2

Keywords

Search

Navigation