doi:

DOI: 10.3724/SP.J.1004.2010.00593

Acta Automatica Sinica (自动化学报) 2010/36:4 PP.593-597

Decentralized Fault Diagnosis of Large-scale Processes Using Multiblock Kernel Principal Component Analysis


Abstract:
In this paper, a multiblock kernel principal com-ponent analysis (MBKPCA) algorithm is proposed. Based on MBKPCA, a new fault detection and diagnosis approach is pro-posed to monitor large-scale processes. In particular, definitions of nonlinear block contributions to T2 and the squared predic-tion error (SPE) statistics are first proposed in order to diagnose nonlinear faults. In addition, the relative contribution, which is the ratio of the contribution to the corresponding upper control limit, is considered to find process variables or blocks responsi-ble for faults. The proposed method is applied to fault detection and diagnosis in the Tennessee Eastman process. The proposed decentralized nonlin-ear approach effectively captures the nonlinear relationship in the block process variables and shows superior fault diagnosis ability compared with other methods.

Key words:Multiblock kernel methods, nonlinear component analysis, process monitoring, fault detection, principal compo-nent analysis

ReleaseDate:2014-07-21 15:17:15

Funds:Supported by National Basic Research Program of China (973 Program) (2009CB320604) and the 111 Project



1 Kano M, Nakagawa Y. Data-based process monitoring, pro-cess control, and quality improvement: recent developments and applications in steel industry. Computers and Chemical Engineering, 2008, 32(1-2): 12-24

2 Yoo C, Kim M, Hwang S, Jo Y, Oh J. Online predictive mon-itoring and prediction model for a periodic process through multiway non-Gaussian modeling. Chinese Journal of Chem-ical Engineering, 2008, 16(1): 48-51

3 Zhao C H, Wang F L, Lu N Y, Jia M X. Stage-based soft-transition multiple PCA modeling and on-line monitoring strategy for batch processes. Journal of Process Control, 2007, 17(9): 728-741

4 Xiong L, Liang J, Qian J X. Multivariate statistical process monitoring of an industrial polypropylene catalyzer reactor with component analysis and kernel density estimation. Chi-nese Journal of Chemical Engineering, 2007, 15(4): 524-532

5 Kruger U, Kumar S, Littler T. Improved principal compo-nent monitoring using the local approach. Automatica, 2007, 43(9): 1532-1542

6 Aguado D, Rosen C. Multivariate statistical monitoring of continuous wastewater treatment plants. Engineering Appli-cations of Artificial Intelligence, 2008, 21(7): 1080-1091

7 Qin S J. Statistical process monitoring: basics and beyond. Journal of Chemometrics, 2003, 17(8-9): 480-502

8 Westerhuis J A, Kourti T, MacGregor J F. Analysis of multi-block and hierarchical PCA and PLS models. Journal of Chemometrics, 1998, 12(5): 301-321

9 Qin S J, Valle S, Piovoso M J. On unifying multiblock anal-ysis with application to decentralized process monitoring. Journal of Chemometrics, 2001, 15(9): 715-742

10 Choi S W, Lee I B. Multiblock PLS--based localized process diagnosis. Journal of Process Control, 2005, 15(3): 295-306

11 MacGregor J F, Jaeckle C, Kiparissides C, Koutoudi M. Pro-cess monitoring and diagnosis by multiblock PLS methods. American Institute of Chemical Engineers Journal, 1994, 40(5): 826-838

12 Kourti T, Nomikos P, MacGregor J F. Analysis, monitor-ing and fault diagnosis of batch processes using multiblock and multiway PLS. Journal of Process Control, 1995, 5(4): 277-284

13 Mark A K. Nonlinear principal component analysis using au-toassociative neural networks. American Institute of Chem-ical Engineers Journal, 1991, 37(2): 233-243

14 Dong D, McAvoy T J. Nonlinear principal component analysis-based on principal curves and neural networks. Computers and Chemical Engineering, 1996, 20(1): 65-78

15 Tan S F, Mavrovouniotis M L. Reducing data dimensionality through optimizing neural network inputs. American Insti-tute of Chemical Engineers Journal, 1995, 41(6): 1471-1480

16 Kocsor A, Toth L. Kernel-based feature extraction with a speech technology application. IEEE Transactions on Signal Processing. 2004, 52(8): 2250-2263

17 Mika S, Scholkopf B, Smola A, Muller K R, Scholz M, Ratsch G. Kernel PCA and de-noising in feature spaces. In: Pro-ceedings on Neural Information Processing Systems. Berlin, UK: Springer, 1991. 536-542

18 Romdhani S, Gong S, Psarrou A. A multi-view nonlin-ear active shape model using kernel PCA. In: Proceed-ings of British Machine Vision Conference. Nottingham, UK: Springer, 1999. 483-492

19 Scholkopf B, Smola, A, Muller K R. Nonlinear component analysis as a kernel eigenvalue problem. Neural Computa-tion, 1998, 10(5): 1299-1319

20 Scholkopf B, Mika S, Burges C J C, Knirsch P, Muller K R, Ratsch G. Input space versus feature space in kernel-based methods. IEEE Transactions on Neural Networks, 1999, 10(5): 1000-1017

21 Zhang Y W, Qin S J. Fault detection of nonlinear processes using multiway kernel independent component analysis. In-dustrial and Engineering Chemistry Research, 2007, 46(23): 7780-7787

22 Zhang Y W, Qin S J. Improved nonlinera fault detection technique and statistical analysis. American Institute of Chemical Engineers Journal, 2008, 54(12): 3207-3220

23 Lee J M, Yoo C K, Lee I B. Fault detection of batch processes using multiway kernel principal component analysis. Com-puters and Chemical Engineering, 2004, 28(9): 1837-1847

24 Lee J M, Yoo C K, Choi S W, Vanrolleghem P A, Lee I B. Nonlinear process monitoring using kernel principal compo-nent analysis. Chemical Engineering Science, 2004, 59(1): 223-234

PDF