Monitoring Nonlinear Batch Process Using Statis-Based Method

Abstract:

Industrial batch processing is widely used in a number of areas of industrial production. Data arising from such processes may present special characteristics; there is therefore a growing interest in the development of customized multivariate control charts for their monitoring. Here we investigate a recent approach that uses control charts based on the Statis method. Statis is an exploratory technique for measuring similarities between data matrices. However, the technique only assesses similarities in a linear context, i.e. investigating structures of linear correlation in the data. In this paper we propose control charts based on the Statis method in conjunction with a kernel for monitoring processes in the presence of nonlinearities. Through kernels we define nonlinear functions of data giving better representation of the structure to be characterized by the Statis method. The new approach is illustrated using simulated data.

• Periodical: Applied Mechanics and Materials (Volume 518)
• Pages: 350-355
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.518.350
• Online since: February 2014
• Authors: D. Marcondes Filho*, F.S. Fogliatto, L.P.L. de Oliveira
• Keywords: Control Charts, Kernels, Nonlinear Batch Processes, Statis Method

References

[1] Nomikos, P.; Macgregor, J.F., Multivariate SPC charts for monitoring batch processes, Technometrics, 1995, 37, 41-59.

DOI: 10.1080/00401706.1995.10485888

[2] Dong, D.; Mcavoy, T.J., Nonlinear principal components analysis based on principal curves and neural networks, Computers and Chemical Engineering, 1996, 20, 65-78.

DOI: 10.1016/0098-1354(95)00003-k

[3] Scholkopf, B.; Smola, A., Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond, MIT Press: Cambridge (MA), (2002).

DOI: 10.1145/1027914.1027921

[4] Lee, J.; Yoo, C. K.; Lee, I. -B., Nonlinear process monitoring using kernel principal component analysis, Chemical Engineering Science, 2004, 59, 223-234.

DOI: 10.1016/j.ces.2003.09.012

[5] Lee, J-M.; Yoo, C.; Choi, S.W.; Vanrolleghem, P.A.; LEE, I-B., Nonlinear process monitoring using kernel principal components analysis, Chemical Engineering Science, 2004, 28, 1837-1847.

DOI: 10.1016/j.ces.2003.09.012

[6] Cui, P.; Li, J.; Wang, G., Improved kernel principal component analysis for fault detection, Expert Systems with Applications, 2008, 34, 1210-1219.

DOI: 10.1016/j.eswa.2006.12.010

[7] Choi, S.W.; Morris, J.; Lee, I. -B., Nonlinear multiscale modeling for fault detection and identification, Chemical Engineering Science, 2008, 63, 2252-2266.

DOI: 10.1016/j.ces.2008.01.022

[8] Zhao, X.Q.; Li, Z.M., An improved KPCA method of fault detection based on wavelet denoising, Key Engeneering Materials, 2011, 467-469, 1427-1432.

DOI: 10.4028/www.scientific.net/kem.467-469.1427

[9] Lavit, C.; Escoufier, Y.; Sabatier, R.; Traissac, P., The ACT (STATIS method), Computational Statistics and Data Analysis, 1994, 19, 97-119.

DOI: 10.1016/0167-9473(94)90134-1

[10] Scepi, G., Parametric and non parametric multivariate quality control charts. In Multivariate Total Quality Control, Physica-Verlang, Lauro C. et al. (eds. ), (2002).

DOI: 10.1007/978-3-642-48710-1_5

[11] Fogliatto, F. S.; Niang, N.; Controle multivariado de processos em batelada com duração variada, Produção, 2008, 18, 240-259.

DOI: 10.1590/s0103-65132008000200004

[12] Boser, B. E.; Guyon, I. M.; Vapnk, V.; 1992, A training algorithm for optimal margin classifiers, In Fifth Annual Workshop on Computational Learning Theory, Pittsburgh, ACM.

DOI: 10.1145/130385.130401

[13] Hastie, T.; Tibshirani, R.; Friedman, J.; The elements of statistical learning – Data mining, inference, and prediction, Springer Science: New York, (2001).

[14] Patel, V. A. A.; Numerical Analysis. New York: Saunders College Publishing, (1993).