变量加权型主元分析算法及其在故障检测中的应用

doi:10.11949/j.issn.0438-1157.20170281

摘要/Abstract

摘要：

传统主成分分析（PCA）算法旨在挖掘训练数据各变量间的相关性特征，已在数据驱动的故障检测领域得到了广泛的研究与应用。然而，传统PCA方法在建模过程中通常认为各个测量变量的重要性是一致的，因此不能有效而全面地描述出变量间相关性的差异。为此，提出一种变量加权型PCA（VWPCA）算法并将之应用于故障检测。首先，通过对训练数据进行加权处理，使处理后的数据能够充分体现出变量间相关性的差异。然后，在此基础上建立分布式的PCA故障检测模型。在线实施故障检测时，则通过贝叶斯准则将多组监测结果融合为一组概率指标。VWPCA方法通过相关性大小为各变量赋予不同的权值，从而将相关性差异考虑进了PCA的建模过程中，相应模型对训练数据特征的描述也就更全面。最后，通过在TE过程上的测试验证VWPCA方法用于故障检测的优越性。

关键词: 主元分析, 过程系统, 过程控制, 故障检测

Abstract:

Traditional principal component analysis (PCA) algorithm, targeting to explore correlations among measured variables in training dataset, has been intensively investigated and applied to data-driven fault detection. However, all variables are considered equally important in modeling process of traditional PCA-based methods, the difference between variable correlations cannot be comprehensively described. A variable weighted PCA (VWPCA) algorithm was proposed and applied to fault detection. Weight calculations were performed on the training dataset so correlation differences among measured variables were fully reflected in the processed data and a distributed PCA-based fault detection model was constructed. When implemented in online fault detection, the Bayesian inference was used to combine multiple monitoring results into an ensemble of probability indices. VWPCA approach assigned different weights to different variables according to the correlation difference, thus PCA modeling took correlation difference into account and the models could completely describe characteristics of the training dataset. Finally, superiority of the proposed VWPCA method was validated by well-known TE process.

Key words: principal component analysis, process systems, process control, fault detection

中图分类号:

TP277

蓝艇, 童楚东, 史旭华. 变量加权型主元分析算法及其在故障检测中的应用[J]. 化工学报, 2017, 68(8): 3177-3182.

LAN Ting, TONG Chudong, SHI Xuhua. Variable weighted principal component analysis algorithm and its application in fault detection[J]. CIESC Journal, 2017, 68(8): 3177-3182.

参考文献 30

[1]	GE Z, SONG Z, GAO F. Review of recent research on data-based process monitoring[J]. Ind. Eng. Chem. Res., 2013, 52(10):3543-3562.
[2]	YIN S, DING S X, XIE X, et al. A review on basic data-driven approaches for industrial process monitoring[J]. IEEE Trans. Ind. Electron., 2014, 61(11):6418-6428.
[3]	周乐, 宋执环, 侯北平, 等. 一种鲁棒半监督建模方法及其在化工过程故障检测中的应用[J]. 化工学报, 2017, 68(3):1109-1115. ZHOU L, SONG Z H, HOU B P, et al. Robust semi-supervised modelling method and its application to fault detection in chemical processes[J]. CIESC Journal, 2017, 68(3):1109-1115.
[4]	王磊, 邓晓刚, 徐莹, 等. 基于变量子域PCA的故障检测方法[J]. 化工学报, 2016, 67(10):4300-4308. WANG L, DENG X G, XU Y, et al. Fault detection method based on variable sub-region PCA[J]. CIESC Journal, 2016, 67(10):4300-4308.
[5]	童楚东, 史旭华. 基于互信息的PCA方法及其在过程监测中的应用[J]. 化工学报, 2015, 66(10):4101-4106. TONG C D, SHI X H. Mutual information based PCA algorithm with application in process monitoring[J]. CIESC Journal, 2015, 66(10):4101-4106.
[6]	江伟, 王振雷, 王昕. 基于混合分块DMICA-PCA的全流程过程监控方法[J]. 化工学报, 2017, 68(2):759-766. JIANG W, WANG Z L, WANG X. Plant-wide process monitoring based on mixed multiblock DMICA-PCA[J]. CIESC Journal, 2017, 68(2):759-766.
[7]	TONG C, LAN T, SHI X. Double-layer ensemble monitoring of non-Gaussian processes using modified independent component analysis[J]. ISA Trans., 2017, 68:181-188.
[8]	胡永兵, 高学金, 李亚芬, 等. 基于仿射传播聚类子集主元分析的间歇过程监测方法[J]. 化工学报, 2016,67(5):1989-1997. HU Y B, GAO X J, LI Y F, et al. Subset multiway principal component analysis monitoring for batch process based on affinity propagation clustering[J]. CIESC Journal, 2016,67(5):1989-1997.
[9]	RATO T J, REIS M S. Fault detection in the Tennessee Eastman benchmark process using dynamic principal components analysis based on decorrelated residuals (DPCA-DR)[J]. Chemom. Intell. Lab. Syst., 2013, 125(7):101-108.
[10]	韩敏,张占奎. 基于改进核主成分分析的故障检测与诊断方法[J]. 化工学报, 2015, 66(6):2139-2149. HAN M, ZHANG Z K. Fault detection and diagnosis method based on modified kernel principal component analysis[J]. CIESC Journal, 2015, 66(6):2139-2149.
[11]	童楚东, 蓝艇, 史旭华. 基于互信息的分散式动态PCA故障检测方法[J]. 化工学报, 2016, 67(10):4317-4323. TONG C D, LAN T, SHI X H. Fault detection by decentralized dynamic PCA algorithm on mutual information[J]. CIESC Journal, 2016, 67(10):4317-4323.
[12]	LIU Y, ZHANG G, XU B. Compressive sparse principal component analysis for process supervisory monitoring and fault detection[J]. J. Process Control, 2017, 50:1-10.
[13]	GE Z, SONG Z. Distributed PCA model for plant-wide process monitoring[J]. Ind. Eng. Chem. Res., 2013, 52(5):1947-1957.
[14]	TONG C, LAN T, SHI X. Fault detection and diagnosis of dynamic processes using weighted dynamic decentralized PCA approach[J]. Chemom. Intell. Lab. Syst., 2017, 161:34-42.
[15]	TONG C, SONG Y, YAN X. Distributed statistical process monitoring based on four-subspace construction and Bayesian inference[J]. Ind. Eng. Chem. Res., 2013, 52(29):9897-9907.
[16]	JIANG Q, WANG B, YAN X. Multiblock independent component analysis integrated with Hellinger distance and Bayesian inference for non-Gaussian plant-wide process monitoring[J]. Ind. Eng. Chem. Res., 2015, 54(9):2497-2508.
[17]	DOWNS J J, VOGEK E F. A plant-wide industrial process control problem[J]. Comput. Chem. Eng., 1993, 17(3):245-255.
[18]	YIN S, DING S X, HAGHANI A, et al. A comparison study of basic data-driven fault diagnosis and process monitoring methods on the benchmark Tennessee Eastman process[J]. J. Process Control, 2012, 22:1567-1581.
[19]	QIN S J. Statistical process monitoring:basics and beyond[J]. J. Chemom., 2003, 17(7/8):480-502.
[20]	GE Z, ZHANG M, SONG Z. Nonlinear process monitoring based on linear subspace and Bayesian inference[J]. J. Process Control, 2010, 20(5):676-688.
[21]	GE Z, SONG Z. Multimode process monitoring based on Bayesian method[J]. J. Chemom., 2009, 23(12):636-650.
[22]	LI N, YANG Y. Ensemble kernel principal component analysis for improved nonlinear process monitoring[J]. Ind. Eng. Chem. Res., 2015, 54(1):318-329.
[23]	GE Z, SONG Z. Bayesian inference and joint probability analysis for batch process monitoring[J]. AIChE J., 2013, 59(10):3702-3713.
[24]	HUANG J, YAN X. Dynamic process fault detection and diagnosis based on dynamic principal component analysis, dynamic independent component analysis and Bayesian inference[J]. Chemom. Intell. Lab. Syst., 2015, 148:115-127.
[25]	杨健, 宋冰, 谭帅, 等. 时序约束NPE算法在化工过程故障检测中的应用[J]. 化工学报, 2016, 67(12):5131-5139. YANG J, SONG B, TAN S, et al. Time constrained NPE for fault detection in chemical processes[J]. CIESC Journal, 2016, 67(12):5131-5139.
[26]	SEVERSON K, CHAIWATANODOM P, BRAATZ R D. Perspectives on process monitoring of industrial systems[J]. Annu. Rev. Control, 2016, 42:190-200.
[27]	ZHANG H, QI Y, WANG L, et al. Fault detection and diagnosis of chemical process using enhanced KECA[J]. Chemom. Intell. Lab. Syst., 2017, 161:61-69.
[28]	BERNAL-DE-LAZARO J M, LLANES-SANTIAGO O, PRIETO-MORENO A, et al. Enhanced dynamic approach to improve the detection of small-magnitude faults[J]. Chem. Eng. Sci., 2016, 14:166-179.
[29]	薄翠梅, 韩晓春, 易辉. 基于聚类选择k近邻的LLE算法及故障检测[J]. 化工学报, 2016, 67(3):925-930. BO C M, HAN X C, YI H, et al. Neighborhood selection of LLE based on cluster for fault detection[J]. CIESC Journal, 2016, 67(3):925-930.et al. Subset multiway principal component analysis monitoring for batch process based on affinity propagation clustering[J]. CIESC Journal, 2016; 67(5):1989-1997.
[9]	RATO T J, REIS M S. Fault detection in the Tennessee Eastman benchmark process using dynamic principal components analysis based on decorrelated residuals (DPCA-DR)[J]. Chemom. Intell. Lab. Syst., 2013, 125(7):101-108.
[10]	Han Min, Zhang Zhankui. Fault detection and diagnosis method based on modified kernel principal component analysis[J]. Journal of Chemical Industry and Engineering (China) (化工学报), 2015; 66(6):2139-2149.
[11]	童楚东, 蓝艇, 史旭华. 基于互信息的分散式动态PCA故障检测方法[J]. 化工学报, 2016, 67(10):4317-4323. TONG C, LAN T, SHI X. Fault detection by decentralized dynamic PCA algorithm on mutual information[J]. CIESC Journal, 2016, 67(10):4317-4323.
[12]	LIU Y, ZHANG G, XU B. Compressive sparse principal component analysis for process supervisory monitoring and fault detection. J. Process Control, 2017, 50:1-10.
[13]	GE Z, SONG Z. Distributed PCA model for plant-wide process monitoring[J]. Ind. Eng. Chem. Res., 2013, 52(5):1947-1957.
[14]	TONG C, LAN T, SHI X. Fault detection and diagnosis of dynamic processes using weighted dynamic decentralized PCA approach[J]. Chemom. Intell. Lab. Syst., 2017, 161:34-42.
[15]	TONG C, SONG Y, YAN X. Distributed Statistical Process Monitoring Based on Four-Subspace Construction and Bayesian Inference[J]. Ind. Eng. Chem. Res., 2013, 52(29):9897-9907.
[16]	JIANG Q, WWANG B, YAN X. Multiblock independent component analysis integrated with Hellinger distance and Bayesian inference for non-Gaussian plant-wide process monitoring[J]. Ind. Eng. Chem. Res., 2015, 54(9):2497-2508.
[17]	DOWNS J J, VOGEK E F. A plant-wide industrial process control problem[J]. Comput. Chem. Eng. 1993, 17(3):245-255.
[18]	YIN S, DING S X, HAGHANI A, HAO H, ZHANG P. A comparison study of basic data-driven fault diagnosis and process monitoring methods on the benchmark Tennessee Eastman process. J. Process Control, 2012, 22:1567-1581.
[19]	RATO T J, REIS M. Fault detection in the Tennessee Eastman benchmark process using dynamic principal components analysis based on decorrelated residuals (DPCA-DR). Chemom. Intell. Lab. Syst., 2013, 125:101-108.
[20]	QIN S J. Statistical process monitoring:basics and beyond[J]. J. Chemom., 2003, 17(7-8):480-502.
[21]	GE Z, ZHANG M, SONG Z. Nonlinear process monitoring based on linear subspace and Bayesian inference[J]. J.Process Control, 2010, 20(5):676-688.
[22]	GE Z, SONG Z. Multimode process monitoring based on Bayesian method[J]. J. Chemom., 2009, 23(12):636-650.
[23]	LI N, YANG Y. Ensemble kernel principal component analysis for improved nonlinear process monitoring[J]. Ind. Eng. Chem. Res., 2015, 54(1):318-329.
[24]	GE Z, SONG Z. Bayesian inference and joint probability analysis for batch process monitoring[J]. AIChE J., 2013, 59(10):3702-3713.
[25]	HUANG J, YAN X. Dynamic process fault detection and diagnosis based on dynamic principal component analysis, dynamic independent component analysis and Bayesian inference[J]. Chemom. Intell. Lab. Syst., 2015, 148:115-127.
[26]	杨健, 宋冰, 谭帅, 等. 时序约束NPE算法在化工过程故障检测中的应用[J]. 化工学报, 2016, 67(12):5131-5139. YANG J, SONG B, TAN S, et al. Time constrained NPE for fault detection in chemical processes[J]. CIESC Journal, 2016, 67(12):5131-5139.
[27]	SEVERSON K, CHAIWATANODOM P, BRAATZ R D. Perspectives on process monitoring of industrial systems[J]. Annu. Rev. Control, 2016, 42:190-200.
[28]	ZHANG H, Qi Y, WANG L, GAO X, WANG X. Fault detection and diagnosis of chemical process using enhanced KECA[J]. Chemom. Intell. Lab. Syst., 2017, 161:61-69.
[29]	BERNAL-DE-LAZARO J M, LLANES-SANTIAGO O, PRIETO-MORENO A, et al. Enhanced dynamic approach to improve the detection of small-magnitude faults[J]. Chem. Eng. Sci., 2016, 14:166-179.
[30]	薄翠梅, 韩晓春, 易辉. 基于聚类选择k近邻的LLE算法及故障检测[J]. 化工学报), 2016, 67(3):925-930. BO C, HAN X, YI H, et al. Neighborhood selection of LLE based on cluster for fault detection[J]. CIESC Journal, 2016, 67(3):925-930.