Similar articles

Abstract: Multiscale classification has potential advantages for monitoring industrial processes generally driven by events in different time and frequency domains. In this study, we adopt stationary wavelet transform for multiscale analysis and propose an applicable scale selection method to obtain the most discriminative scale features. Then using the multiscale features, we construct two classifiers: (1) a supported vector machine (SVM) classifier based on classification distance, and (2) a Bayes classifier based on probability estimation. For the SVM classifier, we use 4-fold cross-validation and grid-search to obtain the optimal parameters. For the Bayes classifier, we introduce dimension reduction techniques including kernel Fisher discriminant analysis (KFDA) and principal component analysis (PCA) to investigate their influence on classification accuracy. We tested the classifiers with two simulated benchmark processes: the continuous stirred tank reactor (CSTR) process and the Tennessee Eastman (TE) process. We also tested them on a real polypropylene production process. The performance comparison among the classifiers in different scales and scale combinations showed that when datasets present typical scale features, the multiscale classifier had higher classification accuracy than conventional single scale classifiers. We also found that dimension reduction can generally contribute to a better classification in our tests.

[1]Aradhye, H.B., Bakshi, B.R., Strauss, R.A., Davis, J.F., 2003. Multiscale SPC using wavelets: theoretical analysis and properties. AIChE J., 49(4):939-958.

[2]Aradhye, H.B., Bakshi, B.R., Davis, J.F., Ahalt, S.C., 2004. Clustering in wavelet domain: a multiresolution ART network for anomaly detection. AIChE J., 50(10):2455-2466.

[3]Bakshi, B.R., 1998. Multiscale PCA with application to multivariate statistical process monitoring. AIChE J., 44(7):1596-1610.

[4]Baudat, G., Anouar, F.E., 2000. Generalized discriminant analysis using a kernel approach. Neur. Comput., 12(10):2385-2404.

[5]Bian, Z., Zhang, X., 2000. Pattern Recognition (2nd Ed.). Tsinghua University Press, Beijing, p.298-299 (in Chinese).

[6]Chiang, L.H., Russell, E.L., Braatz, R.D., 2000. Fault diagnosis in chemical processes using Fisher discriminant analysis, discriminant partial least squares, and principal component analysis. Chemometr. Intell. Lab. Syst., 50(2):243-252.

[7]Chiang, L.H., Kotanchek, M.E., Kordon, A.K., 2004. Fault diagnosis based on Fisher discriminant analysis and support vector machines. Comput. Chem. Eng., 28(8):1389-1401.

[8]Detroja, K.P., Gudi, R.D., Patwardhan, S.C., 2006. A possibilistic clustering approach to novel fault detection and isolation. J. Process Control, 16(10):1055-1073.

[9]Downs, J.J., Vogel, E.F., 1993. A plant-wide industrial process control problem. Comput. Chem. Eng., 17(3):245-255.

[10]He, Q.P., Qin, S.J., Wang, J., 2005. A new fault diagnosis method using fault directions in Fisher discriminant analysis. AIChE J., 51(2):555-571.

[11]He, X.B., Yang, Y.P., Yang, Y.H., 2008. Fault diagnosis based on variable-weighted kernel Fisher discriminant analysis. Chemometr. Intell. Lab. Syst., 93(1):27-33.

[12]Hsu, C.W., Lin, C.J., 2002. A comparison of methods for multiclass support vector machines. IEEE Trans. Neur. Networks, 13(2):415-425.

[13]Hsu, C.W., Chang, C.C., Lin, C.J., 2008. A Practical Guide to Support Vector Classification. Available from http://www.csie.ntu.edu.tw/~cjlin/papers/guide/guide.pdf [Accessed on May 21, 2010].

[14]Li, J., Zhang, J., Ge, W., Liu, X., 2004. Multi-scale methodology for complex systems. Chem. Eng. Sci., 59(8-9):1687-1700.

[15]Misra, M., Yue, H.H., Qin, S.J., Ling, C., 2002. Multivariate process monitoring and fault diagnosis by multi-scale PCA. Comput. Chem. Eng., 26(9):1281-1293.

[16]Percival, D.B., Walden, A.T., 2000. Wavelet Methods for Time Series Analysis. Cambridge University Press, Cambridge, p.160, 195-200.

[17]Reis, M.S., Bauer, A., 2009. Wavelet texture analysis of on-line acquired images for paper formation assessment and monitoring. Chemometr. Intell. Lab. Syst., 95(2):129-137.

[18]Reis, M.S., Saraiva, P.M., 2006. Generalized multiresolution decomposition frameworks for the analysis of industrial data with uncertainty and missing values. Ind. Eng. Chem. Res., 45(18):6330-6338.

[19]Reis, M.S., Saraiva, P.M., Bakshi, B.R., 2008. Multiscale statistical process control using wavelet packets. AIChE J., 54(9):2366-2378.

[20]Russell, E.L., Chiang, L.H., Braatz, R.D., 2000. Data-Driven Methods for Fault Detection and Diagnosis in Chemical Process. Springer, London, p.64, 103-107.

[21]Wang, H., Li, P., Gao, F., Song, Z., Ding, S.X., 2006. Kernel classifier with adaptive structure and fixed memory for process diagnosis. AIChE J., 52(10):3515-3531.

[22]Woody, A.A., Brown, S.D., 2007. Selecting wavelet transform scales for multivariate classification. J. Chemometr., 21(7-9):357-363.

[23]Yoon, S., MacGregor, J.F., 2001. Fault diagnosis with multivariate statistical models: part I. using steady state fault signatures. J. Process Control, 11(4):387-400.

[24]Yoon, S., MacGregor, J.F., 2004. Principal-component analysis of multiscale data for process monitoring and fault diagnosis. AIChE J., 50(11):2891-2903.

[25]Yu, J., Qin, S.J., 2008. Multimode process monitoring with Bayesian inference-based finite Gaussian mixture models. AIChE J., 54(7):1811-1829.

[26]Zhou, S., Xie, L., Wang, S., 2005. On-line fault diagnosis in industrial process using variable moving window and hidden Markov model. Chin. J. Chem. Eng., 13(3):388-395.