Comparison of linear and nonlinear dimension reduction techniques for automated process monitoring of a decentralized wastewater treatment facility

  • Karen Kazor
  • Ryan W. Holloway
  • Tzahi Y. Cath
  • Amanda S. Hering
Original Paper

Abstract

Multivariate statistical methods for online process monitoring have been widely applied to chemical, biological, and engineered systems. While methods based on principal component analysis (PCA) are popular, more recently kernel PCA (KPCA) and locally linear embedding (LLE) have been utilized to better model nonlinear process data. Additionally, various forms of dynamic and adaptive monitoring schemes have been proposed to address time-varying features in these processes. In this analysis, we extend a common simulation study in order to account for autocorrelation and nonstationarity in process data and comprehensively compare the monitoring performances of static, dynamic, adaptive, and adaptive–dynamic versions of PCA, KPCA, and LLE. Furthermore, we evaluate a nonparametric method to set thresholds for monitoring statistics and compare results with the standard parametric approaches. We then apply these methods to real-world data collected from a decentralized wastewater treatment system during normal and abnormal operations. From the simulation study, adaptive–dynamic versions of all three methods generally improve results when the process is autocorrelated and nonstationary. In the case study, adaptive–dynamic versions of PCA, KPCA, and LLE all flag a strong system fault, but nonparametric thresholds considerably reduce the number of false alarms for all three methods under normal operating conditions.

Keywords

Multivariate statistical process control Nonlinear time-varying process Principle component analysis Kernel principal component analysis Locally linear embedding 

Supplementary material

477_2016_1246_MOESM1_ESM.pdf (102 kb)
Supplementary material 1 (pdf 102 KB)

References

  1. Banadda N, Nhapi I, Kimwaga R (2011) A review of modeling approaches in activated sludge systems. Afr J Environ Sci Technol 5:397–408Google Scholar
  2. Belkin M, Niyogi P (2003) Laplacian eigenmaps for dimensionality reduction and data representation. Neural Comput 15:1373–1396CrossRefGoogle Scholar
  3. Chen Q, Wynne RJ, Goulding P, Sandoz D (2000) The application of principal component analysis and kernel density estimation to enhance process monitoring. Control Eng Pract 8:531–543CrossRefGoogle Scholar
  4. Cheng C, Chiu MS (2005) Nonlinear process monitoring using JITL-PCA. Chemom Intell Lab Syst 76:1–13CrossRefGoogle Scholar
  5. Cheng CY, Hsu CC, Chen MC (2010) Adaptive kernel principal component analysis (KPCA) for monitoring small disturbances of nonlinear processes. Ind Eng Chem Res 49:2254–2262CrossRefGoogle Scholar
  6. Chetouani Y (2008) A neural network approach for the real-time detection of faults. Stoch Environ Res Risk Assess 22:339–349CrossRefGoogle Scholar
  7. Cho JH, Lee JM, Choi SW, Lee D, Lee IB (2005) Fault identification for process monitoring using kernel principal component analysis. Chem Eng Sci 60:279–288CrossRefGoogle Scholar
  8. Choi SW, Lee IB (2004) Nonlinear dynamic process monitoring based on dynamic kernel PCA. Chem Eng Sci 59:5897–5908CrossRefGoogle Scholar
  9. Choi SW, Lee C, Lee JM, Park JH, Lee IB (2005) Fault detection and identification of nonlinear processes based on kernel PCA. Chemom Intell Lab Syst 75:55–67CrossRefGoogle Scholar
  10. Chouaib C, Mohamed-Faouzi H, Messaoud D (2013) Adaptive kernel principal component analysis for nonlinear dynamic process monitoring. In: 2013 9th Asian Control Conference (ASCC). IEEE, Istanbul, TurkeyGoogle Scholar
  11. Dong D, McAvoy TJ (1996) Batch tracking via nonlinear principal component analysis. AIChE J 42:2199–2208CrossRefGoogle Scholar
  12. Downs JJ, Vogel EF (1993) A plant-wide industrial process control problem. Comput Chem Eng 17:245–255CrossRefGoogle Scholar
  13. García-Díaz CJ (2011) Monitoring and forecasting nitrate concentration in the groundwater using statistical process control and time series analysis: a case study. Stoch Environ Res Risk Assess 25:331–339CrossRefGoogle Scholar
  14. Ge Z, Yang C, Song Z (2009) Improved kernel PCA-based monitoring approach for nonlinear processes. Chem Eng Sci 64:2245–2255CrossRefGoogle Scholar
  15. Gikas P, Tchobanoglous G (2009) The role of satellite and decentralized strategies in water resources management. J Environ Manag 90:144–152CrossRefGoogle Scholar
  16. Ham J, Lee DD, Mika S, Sch\(\ddot{\text{o}}\)lkopf B (2004) A kernel view of the dimensionality reduction of manifolds. In: Proceedings of the Twenty-First International Conference on Machine Learning. ACM, New York, USA, p 47Google Scholar
  17. Hart JD (1996) Some automated methods of smoothing time-dependent data. J Nonparametr Stat 6:115–142CrossRefGoogle Scholar
  18. Haykin S (1999) Neural Networks. Prentice-Hall, Englewood CliffsGoogle Scholar
  19. Hiden HG, Willis MJ, Tham MT, Montague GA (1999) Non-linear principal components analysis using genetic programming. Comput Chem Eng 23:413–425CrossRefGoogle Scholar
  20. Jackson JE, Mudholkar GS (1979) Control procedures for residuals associated with principal component analysis. Technometrics 21:341–349CrossRefGoogle Scholar
  21. Jones MC, Marron JS, Sheather SJ (1996) A brief survey of bandwidth selection for density estimation. J Am Stat Assoc 91:401–407CrossRefGoogle Scholar
  22. Kayo O (2006) Locally linear embedding algorithm: extensions and applications. Universitatis Ouluensis, Oulu. http://herkules.oulu.fi/isbn9514280415/isbn9514280415.pdf. Accessed 18 March 2014
  23. Khediri IB, Limam M, Weihs C (2011) Variable window adaptive kernel principal component analysis for nonlinear nonstationary process monitoring. Comput Ind Eng 61:437–446CrossRefGoogle Scholar
  24. Kramer MA (1991) Nonlinear principal component analysis using autoassociative neural networks. AIChE J 37:233–243CrossRefGoogle Scholar
  25. Kruger U, Zhou Y, Irwin GW (2004) Improved principal component monitoring of large-scale processes. J Process Control 14:879–888CrossRefGoogle Scholar
  26. Kruger U, Antory D, Hahn J, Irwin GW, McCullough G (2005) Introduction of a nonlinearity measure for principal component models. Comput Chem Eng 29:2355–2362CrossRefGoogle Scholar
  27. Ku W, Storer RH, Georgakis C (1995) Disturbance detection and isolation by dynamic principal component analysis. Chemom Intell Lab Syst 30:179–196CrossRefGoogle Scholar
  28. Lee JM, Yoo CK, Choi SW, Vanrolleghem PA, Lee IB (2004a) Nonlinear process monitoring using kernel principal component analysis. Chem Eng Sci 59:223–234CrossRefGoogle Scholar
  29. Lee JM, Yoo CK, Lee IB (2004b) Statistical process monitoring with independent component analysis. J Process Control 14:467–485CrossRefGoogle Scholar
  30. Leverenz HL, Asano A (2011) Wastewater reclamation and reuse system. Treatise Water Sci 4:63–71CrossRefGoogle Scholar
  31. Liu X, Kruger U, Littler T, Xie L, Wang S (2009) Moving window kernel PCA for adaptive monitoring of nonlinear processes. Chemom Intell Lab Syst 96:132–143CrossRefGoogle Scholar
  32. Maulud A, Wang D, Romagnoli JA (2006) A multi-scale orthogonal nonlinear strategy for multivariate statistical process monitoring. J Process Control 16:671–683CrossRefGoogle Scholar
  33. Miao A, Song Z, Ge Z, Zhou L, Wen Q (2013) Nonlinear fault detection based on locally linear embedding. J Control Theory Appl 11:615–622CrossRefGoogle Scholar
  34. Mina J, Verde C (2007) Fault detection for large scale systems using Dynamic Principal Components Analysis with adaptation. Int J Comput Commun Control 2:185–194CrossRefGoogle Scholar
  35. Mjalli FS, Al-Asheh S, Alfadala HE (2007) Use of artificial neural network black-box modeling for the prediction of wastewater treatment plants performance. J Environ Manag 83:329–338CrossRefGoogle Scholar
  36. Nguyen VH, Golinval JC (2010) Fault detection based on kernel principal component analysis. Eng Struct 32:3683–3691CrossRefGoogle Scholar
  37. Parneet P (2013) Comparing and contrasting traditional membrane bioreactor models with novel ones based on time series analysis. Membranes 3:16–23CrossRefGoogle Scholar
  38. Rato T, Reis M (2013) Defining the structure of DPCA models and its impact on process monitoring and prediction activities. Chemom Intell Lab Syst 125:74–86CrossRefGoogle Scholar
  39. Roweis ST, Saul LK (2000) Nonlinear dimensionality reduction by locally linear embedding. Science 290:2323–2326CrossRefGoogle Scholar
  40. Russell EL, Chiang LH, Braatz RD (2000) Fault detection in industrial processes using canonical variate analysis and dynamic principal component analysis. Chemom Intell Lab Syst 51:81–93CrossRefGoogle Scholar
  41. Schölkopf B, Smola A, Müller KB (1998) Nonlinear component analysis as a kernel eigenvalue problem. Neural Comput 10:1299–1319CrossRefGoogle Scholar
  42. Sheather SJ (2009) A modern approach to regression with R, vol 58. Springer, New YorkCrossRefGoogle Scholar
  43. Sheather SJ, Jones MC (1991) A reliable data-based bandwidth selection method for kernel density estimation. J R Stat Soc B 53:683–690Google Scholar
  44. Sköld M (2001) A bias correction for cross-validation bandwidth selection when a kernel estimate is based on dependent data. J Time Ser Anal 22:493–503CrossRefGoogle Scholar
  45. Tenenbaum JB, De Silva V, Langford JC (2000) A global geometric framework for nonlinear dimensionality reduction. Science 290:2319–2323CrossRefGoogle Scholar
  46. van Sprang ENM, Ramaker HJ, Westerhuis JA, Gurden SP, Smilde AK (2002) Critical evaluation of approaches for on-line batch process monitoring. Chem Eng Sci 57:3979–3991CrossRefGoogle Scholar
  47. Vuono D, Henkel J, Benecke J, Cath TY, Reid T, Johnson L, Drewes JE (2013) Flexible hybrid membrane treatment systems for tailored nutrient management: a new paradigm in urban wastewater treatment. J Membr Sci 446:34–41CrossRefGoogle Scholar
  48. Westerhuis JA, Gurden SP, Smilde AK (2000) Generalized contribution plots in multivariate statistical process monitoring. Chemom Intell Lab Syst 51:95–114CrossRefGoogle Scholar
  49. Wold S, Esbensen K, Geladi P (1987) Principal component analysis. Chemom Intell Lab Syst 2:37–52CrossRefGoogle Scholar
  50. Woodall WH, Montgomery DC (2014) Some current directions in the theory and application of statistical process monitoring. J Qual Technol 46:78–94Google Scholar
  51. Zeng Y, Zhang Z, Kusiak A, Tang F, Wei X (2016) Optimizing wastewater pumping system with data-driven models and a greedy electromagnetism-like algorithm. Stoch Environ Res Risk Assess 30:1263–1275CrossRefGoogle Scholar
  52. Zhang W, Liu X, Qi R, Jiang Y (2013) Improved locally linear embedding based method for nonlinear system fault detection. Int J Adv Comput Technol 5:515–523Google Scholar
  53. Zhao S, Xu Y (2005) Multivariate statistical process monitoring using robust nonlinear principal component analysis. Tsinghua Sci Technol 10:582–586CrossRefGoogle Scholar
  54. Zheng Y, Zhang L (2013) Fault diagnosis of wet flue gas desulphurization system based on KPCA. In: The 19th International Conference on Industrial Engineering and Engineering Management. Springer, BerlinGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2016

Authors and Affiliations

  • Karen Kazor
    • 1
  • Ryan W. Holloway
    • 2
  • Tzahi Y. Cath
    • 2
  • Amanda S. Hering
    • 1
  1. 1.Department of Applied Mathematics and StatisticsColorado School of MinesGoldenUSA
  2. 2.Department of Civil and Environmental EngineeringColorado School of MinesGoldenUSA

Personalised recommendations