基于支持向量机MPLS的间歇过程故障诊断方法

• SYSTEM ENGINEERING • 上一篇下一篇

基于支持向量机MPLS的间歇过程故障诊断方法

李运锋^a; 汪志锋^b; 袁景淇^a,c

^aDepartment of Automation, Shanghai Jiao Tong University, Shanghai 200030, China
^b Department of Automation, Shanghai Second Polytechnic University, Shanghai 201209, China
^c State Key Laboratory of Bioreactor Engineering, East China University of Science and Technology, Shanghai 200237, China

收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2006-12-28 发布日期:2006-12-28
通讯作者: 李运锋

On-line fault detection using SVM-based dynamic MPLS for batch processes

LI Yunfeng^a; WANG Zhifeng^b; YUAN Jingqi^a,c

^a Department of Automation, Shanghai Jiao Tong University, Shanghai 200030, China
^b Department of Automation, Shanghai Second Polytechnic University, Shanghai 201209, China
^c State Key Laboratory of Bioreactor Engineering, East China University of Science and Technology, Shanghai 200237, China

Received:1900-01-01 Revised:1900-01-01 Online:2006-12-28 Published:2006-12-28
Contact: LI Yunfeng

摘要/Abstract

摘要： In this article, a nonlinear dynamic multiway partial least squares (MPLS) based on support vector machines (SVM) is developed for on-line fault detection in batch processes. The approach, referred to as SVM-based DMPLS, integrates the SVM with the MPLS model. Process data from normal historical batches are used to develop the MPLS model, and a series of single-input-single-output SVM networks are adopted to approximate nonlinear inner relationship between input and output variables. In addition, the application of a time-lagged window technique not only makes the complementarities of unmeasured data of the monitored batch unnecessary, but also significantly reduces the computation and storage requirements in comparison with the traditional MPLS. The proposed approach is validated by a simulation study of on-line fault detection for a fed-batch penicillin production.

关键词: fault detection;multiway partial least squares;support vector machines;time-lagged window

Abstract: In this article, a nonlinear dynamic multiway partial least squares (MPLS) based on support vector machines (SVM) is developed for on-line fault detection in batch processes. The approach, referred to as SVM-based DMPLS, integrates the SVM with the MPLS model. Process data from normal historical batches are used to develop the MPLS model, and a series of single-input-single-output SVM networks are adopted to approximate nonlinear inner relationship between input and output variables. In addition, the application of a time-lagged window technique not only makes the complementarities of unmeasured data of the monitored batch unnecessary, but also significantly reduces the computation and storage requirements in comparison with the traditional MPLS. The proposed approach is validated by a simulation study of on-line fault detection for a fed-batch penicillin production.

Key words: fault detection, multiway partial least squares, support vector machines, time-lagged window

李运锋,汪志锋, 袁景淇. 基于支持向量机MPLS的间歇过程故障诊断方法[J]. , DOI: .

LIYunfeng,WANGZhifeng, YUANJingqi. On-line fault detection using SVM-based dynamic MPLS for batch processes[J]. , DOI: .

参考文献

1    Golobic, I., Gjerkes, H., Bajsic, I., Malensek, J., “Soft-ware sensor for biomass concentration monitoring during industrial fermentation”, Instrum. Sci. Technol., 28(4), 323—334(2000).
2     Li, Y., Zhou, D.H., Xie, Z., Qin, S.J., “Faulty sensor de-tection and reconstruction for a PVC making process”, Chin. J. Chem. Eng., 12(2), 227—233(2004).
3    Nomikos, P., MacGregor, J. F., “Monitoring batch proc-esses using multiway principal component analysis”, AIChE J., 40, 1361—1375(1994).
4    Song, K., Wang, H.Q., Li, P., “VPLS based quality and cost control for Tennessee Eastman process”, Chin. J. Chem. Eng., 13(1), 62—67(2005).
5    Gregersen, L., Jorgensen, S.B., “Supervision of fed-batch fermentations”, Chem. Eng. J., 75(1), 69—76(1999).
6    Undey, C., Tatara, E., Cinar, A., “Intelligent real-time performance monitoring and quality prediction for batch/fed-batch cultivations”, J. Biotechnol., 108(1),   61—77(2004).
7    Wold, S., Kettaneh-Wold, N., Skagerberg, B., “Nonlin-ear PLS modeling”, Chemom. Intell. Lab. Syst., 7, 53—65(1989).
8    Qin, S.J., McAvoy, T.J., “Nonlinear PLS modeling using neural networks”, Comput. Chem. Eng., 16(4), 379—391 (1992).
9    Vapnik, V.N., The Nature of Statistical Learning Theory, Springer, Berlin (1995).
10    Cai, C.Z., Wang, W.L., Sun, L.Z., Chen, Y.Z., “Protein function classification via support vector machine ap-proach”, Math. Biosci., 185(2), 111—122(2003).
11    Li, Y.F., Yuan, J.Q., “Prediction of key state variables using support vector machines in bioprocesses”, Chem. Eng. Technol., 29(3), 313—319(2006).
12    Cho, H., Kim, K., “A method for predicting future ob-servations in the monitoring of a batch process”, J. Qual. Technol., 57, 63—75(2003).
13    Meng, X., Morris, A. J., Martin, E. B. “On-line monitor-ing of batch processes using a PARAFAC representa-tion”, J. Chemometr., 17, 65—81(2003).
14    Chen, J.H., Liu, K.C., “On-line batch process monitoring using dynamic PCA and dynamic PLS models”, Chem. Eng. Sci., 57(1), 63—75(2002).
15    Ku, W.F., Storer, R.H., Gerorgakis, C., “Disturbance detec-tion and isolation by dynamic principal component analy-sis”, Chemometr. Intell. Lab., 30(1), 179—196(1995).
16    Wold, H., “Nonlinear estimation by iterative least squares procedures”, In: Research Papers in Statistics, David, F., Ed., Wiley, New York (1996).
17    Birol, G., Cerk, U., Ali, C., “A modular simulation package for fed-batch fermentation penicillin produc-tion”, Comput. Chem. Eng., 26, 1553—1565(2002).
18    Cenk, Ü., Eric, T., Ali, C., “Intelligent real-time perform-ance monitoring and quality prediction for batch/fed-batch cultivations”, J. Biotechnol., 108(1), 61—77(2004).