Fault Detection Based on Multivariate Trajectory Analysis

被引：1

作者：

Shen F. ^{[1
]}

Song Z. ^{[1
]}

Ge Z. ^{[1
]}

机构：

[1] State Key Laboratory of Industrial Control Technology, Zhejiang University, Hangzhou

来源：

Hsi-An Chiao Tung Ta Hsueh/Journal of Xi'an Jiaotong University | 2017年 / 51卷 / 03期

关键词：

Fault detection; Multivariate trajectory analysis; Principal component analysis; Synthetic ammonia production;

D O I：

10.7652/xjtuxb201703021

中图分类号：

学科分类号：

摘要：

To solve dynamic problems in industrial multivariable process monitoring, a novel fault detection method is proposed to describe process trajectories combining multivariate trajectory analysis with principal component analysis. The trajectory vectors are constructed to extract information in multivariate dynamic process, and then principal component analysis algorithm is adopted to develop the model and analyze the variation features of process data. The trajectory tendency charts of several critical variables involved with data variations are plotted, and offline modeling and online fault detection are finally realized. Compared with the traditional methods based on trajectory analysis, the proposed method breaks the limitation of variable number and solves the difficulty in developing monitoring statistics to better extract characters in process dynamics and more reliable dynamic process monitoring. A practical case of synthetic ammonia conversion unit verifies the effectiveness of the proposed method. © 2017, Editorial Office of Journal of Xi'an Jiaotong University. All right reserved.

引用

页码：122 / 128

页数：6

共 15 条

[1]

Qin S.J., Survey on data-driven industrial process monitoring and diagnosis, Annual Reviews in Control, 36, 2, pp. 220-234, (2012)

[2]

Tong C., El-Farra N.H., Palazoglu A., Et al., Fault detection and isolation in hybrid process systems using a combined data-driven and observer-design methodology, AIChE Journal, 60, 8, pp. 2805-2814, (2014)

[3]

Ge Z., Song Z., Gao F., Review of recent research on data-based process monitoring, Industrial & Engineering Chemistry Research, 52, 10, pp. 3543-3562, (2013)

[4]

Ge Z., Song Z., Semiconductor manufacturing process monitoring based on adaptive substatistical PCA, IEEE Transactions on Semiconductor Manufacturing, 23, 1, pp. 99-108, (2010)

[5]

Nomikos P., Macgregor J.F., Monitoring batch processes using multiway principal component analysis, AIChE Journal, 40, 8, pp. 1361-1375, (1994)

[6]

Nomikos P., Macgregor J.F., Multi-way partial least squares in monitoring batch processes, Chemometrics and Intelligent Laboratory Systems, 30, 1, pp. 97-108, (1995)

[7]

Wen Q., Ge Z., Song Z., Data-based linear Gaussian state-space model for dynamic process monitoring, AIChE Journal, 58, 12, pp. 3763-3776, (2012)

[8]

Choi S.W., Morris J., Lee I.B., Dynamic model-based batch process monitoring, Chemical Engineering Science, 63, 3, pp. 622-636, (2008)

[9]

Yu J., Multiway discrete hidden Markov model-based approach for dynamic batch process monitoring and fault classification, AIChE Journal, 58, 9, pp. 2714-2725, (2012)

[10]

Bogomolov A., Multivariate process trajectories: capture, resolution and analysis, Chemometrics and Intelligent Laboratory Systems, 108, 1, pp. 49-63, (2011)

← 1 2 →