Fault Detection by Integrating Canonical Variate Analysis and Independent Component Analysis Based on Local Outlier Factor
Subject Areas : electrical and computer engineeringE. Tavasolipour 1 , M. T. Hamidi Beheshti 2 , A. Ramezani 3
1 - Tarbiat Modares University
2 - Tarbiat Modares University
3 - دانشگاه تربیت مدرس
Abstract :
In this paper a novel process monitoring scheme is proposed because of the importance of fault detection and identification in industrial processes. In this method, process dynamic and effect of outliers are considered concurrently. First, the proposed approach uses CVA method to implement the process dynamic. Then ICA method is performed for dimension reduction of data. The outliers elimination and control limit calculation are based on the Local Outlier Factor algorithm. This algorithm doesn’t consider a special distribution for process variables, thus conforming to data in real industrial processes. The proposed method is applied to fault detection in the Tennessee Eastman process. Results clearly indicate better performance of the proposed scheme compared to the alternative methods.
[1] J. V. Kresta, J. F. Macgregor, and T. H. E. Marlin, "Multivariate statistacial monitoring of process operating performance," Can. J. Chem. Eng., vol. 69, no. 1, pp. 35-47, 1991.
[2] B. R. Bakhshi, "Multiscale PCA with application to multivariate statistical process monitoring," AlChE Journal, vol. 44, no. 7, pp. 1596-1610, 1998.
[3] W. Li, H. Yue, S. V. Cervantes, and J. Qin, "Recursive PCA for adaptive process monitoring," J. Process Control, vol. 10, no. 5, pp. 471-486, 2000.
[4] B. M. Wise and N. B. Gallagher, "The process chemometrics approach to process monitoring and fault detection," Comp. Chem. Eng, vol. 20, no. 1, pp. 65-78, 1996.
[5] D. Dong and T. J. MacAvoy, "Nonlinear principal component analysis based on principal curves and neural networks," J. Process Control, vol. 6, no. 6, pp. 329-348, Jan. 1996.
[6] P. Nomikos and J. F. MacGrejor, "Monitoring batch processes using multiway principal component analysis," AlChE Journal, vol. 40, no. 8, pp. 1361-1375, Aug. 1994.
[7] J. M. Lee, C. Yoo, S. W. Choi, P. A. Vanrolleghem, and I. B. Lee, "Nonlinear process monitoring using kernel principal component analysis," Chem. Eng. Sci, vol. 59, pp. 223-234, 2004.
[8] S. W. Choi, C. Lee, J. M. Lee, J. H. Park, and I. B. Lee, "Fault detection and identification of nonlinear process based on kernel PCA," Chemometr. Intell. Lab, vol. 75, no. 1, pp. 55-67, 2005.
[9] A. Hyvarinen and E. Oja, "Independent component analysis: algorithm and applications," Neural Network Control, vol. 13, no. 4-5, pp. 411-413, May/Jun. 2000.
[10] C. Hsu, M. Chen, and L. Chen, "Intelligent ICA-SVM fault detector for non-gaussian multivariate process monitoring," Expert Syst., vol. 37, no. 4, pp. 3264-3273, Apr. 2010.
[11] W. Li and S. Hongbo, "Multivariate statistical process monitoring using an improved independent component analysis," Chemical Engineering Research and Design, vol. 88, no. 4, pp. 403-414, Apr. 2010.
[12] Y. Zangh, "Enhanced statistical analysis of nonlinear using KPCA, KICA and SVM," Chem. Eng. Sci, vol. 64, no. 5, pp. 801-811, Mar. 2009.
[13] Z. Ge and Z. Song, "Process monitoring based on independent component analysis-principal component analysis (ICA-PCA) and similarity factors," Ind. Eng. Chem. Res., vol. 46, no. 7, pp. 2054-2063, Mar. 2007.
[14] X. Liu, L. Xie, U. Kruger, T. Littler, and S. Wang, "Statistical-based monitoring of multivariate non-gaussian systems," AlChE J., vol. 54, pp. 2379-2391, 2008.
[15] R. Storer, W. Ku, and C. Georgakis, "Disturbance detection and isolation by dynamic principal component analysis," Chemometr. Intell. Lab, vol. 30, no. 1, pp. 179-196, Nov. 1995.
[16] M. L. Jong, Y. ChangKyoo, and L. In-Beum, "Statistical monitoring of dynamic processes based on dynamic independent component analysis," Chem. Eng. Sci, vol. 59, no. 14, pp. 2995-3006, Jul. 2004.
[17] M. Isaac, B. Raul, E. Gerard, and G. Moises, "DICA enhanced SVM classification approach to fault diagnosis for chemical processes," in Proc. 19th European Symposium on Computer Aided Process Engineering, ESCAPE'09, vol. 26, pp. 267-272, 2009.
[18] H. Chun-Chin, C. Long-Sheng, and L. Cheng-Hsiang, "A process monitoring scheme based on independent component analysis and adjusted outliers," Int. J. Prod. Res, vol. 48, no. 15, pp. 1727-1743, 2010.
[19] H. Chun-Chin, C. Mu-Chen, and C. Long-Sheng, "A novel process monitoring approach with dynamic independent component analysis," Control Eng. Pract., vol. 18, no. 3, pp. 242-253, Mar. 2010.
[20] J. Lee, B. Kang, and S. H. Kang, "Integrating independent component analysis and local outlier factor for plant-wide process monitoring," J. Process Control, vol. 21, no. 7, pp. 1011-1021, Aug. 2011.
[21] A. Norvalis, A. Negiz, J. DeCicco, and A. Cinar, "Intelligent process monitoring by interfacing knowledge-based systems and multivariate statistical monitoring," J. Process Control, vol. 10, no. 4, pp. 341-350, Aug. 2000.
[22] Y. Yang, Y. Chen, X. Chen, and X. Liu, "Multivariate industrial process monitoring based on the integration method of canonical variate analysis and independent component analysis," Chemometr. Intell. Lab., vol. 116, pp. 94-101, Jul. 2012.
[23] M. M. Breunig, H. P. Kriegel, R. T. Ng, and J. Sander, "Identifying density-based local outliers," in Proc. ACM Int. Conf. on Management of Data, SIGMOD'00, pp. 93-104, 2000.
[24] P. P. Odiowei and Y. Cao, "Nonlinear dynamic process monitoring using canonical variate analysis and kernel density estimations," IEEE Trans. on Industrial Informatics, vol. 6, no. 1, pp. 36-44, Mar. 2010.
[25] L. H. Chiang, E. L. Russell, and R. D. Braatz, Fault Detection and Diagnosis in Industrial Systems, Springer, Urbana-champaing, London, Ch. 1, pp. 103-108.
