Fault detection and diagnosis method based on weighted statistical feature KICA

doi:10.11949/0438-1157.20211295

Abstract

Abstract:

Aiming at the problem of low early fault detection rate in the nonlinear dynamic process of kernel independent component analysis (KICA), a fault detection and diagnosis approach based on weighted statistical feature KICA (WSFKICA) is proposed. First, the independent component data and residual data are captured from the original data by using KICA. Then, the improved statistical feature data set is obtained by weighted statistical feature and sliding window, and the statistics is constructed from the data set for fault detection. Finally, the method based on contribution chart of monitored variables is used for process fault diagnosis. Compared with traditional KICA statistics, the statistics of the proposed approach has higher fault detection performance for incipient faults in nonlinear dynamic processes. The proposed approach is tested in a simulated case and in the Tennessee-Eastman (TE) process. The simulation results show that the proposed approach has an advantage over ICA, KICA, KPCA and statistical local kernel principal component analysis (SLKPCA).

Key words: independent component analysis, incipient faults, statistical feature, process control, parameter estimation, fault diagnosis, dynamic simulation

摘要：

针对核独立元分析（kernel independent component analysis， KICA）在非线性动态过程中对微小故障检测率低的问题，提出一种基于加权统计特征KICA（weighted statistical feature KICA， WSFKICA）的故障检测与诊断方法。首先，利用KICA从原始数据中捕获独立元数据和残差数据；然后，通过加权统计特征和滑动窗口获取改进统计特征数据集，并由此数据集构建统计量进行故障检测；最后，利用基于变量贡献图的方法进行过程故障诊断。与传统KICA统计量相比，所提方法的统计量对非线性动态过程中的微小故障具有更高的故障检测性能。应用该方法对一个数值例子和田纳西-伊斯曼（Tennessee-Eastman， TE）过程进行仿真测试，仿真结果显示出所提方法相对于独立元分析（ICA）、KICA、核主成分分析（kernel principal component analysis， KPCA）和统计局部核主成分分析（statistical local kernel principal component analysis， SLKPCA）检测的优势。

关键词: 独立元分析, 微小故障, 统计特征, 过程控制, 参数估值, 故障诊断, 动态仿真

CLC Number:

TP 277

Cheng ZHANG, Lizhi PAN, Yuan LI. Fault detection and diagnosis method based on weighted statistical feature KICA[J]. CIESC Journal, 2022, 73(2): 827-837.

张成, 潘立志, 李元. 基于加权统计特征KICA的故障检测与诊断方法[J]. 化工学报, 2022, 73(2): 827-837.

Figures/Tables 25

Fig.1 Independent component using KICA

Fig.2 Flow diagram of WSFKICA

Table 1 Validation accuracy at different parameter β

$β$	校验准确率		平均值
$β$	R	H	平均值
$2$	90.8	76.7	83.7
$3$	91.7	75.6	83.6
$4$	97.6	71.4	84.5
$5$	97.6	69.1	83.3
$6$	97.6	65.7	81.6

Table 1 Validation accuracy at different parameter β

$β$	校验准确率		平均值
$β$	R	H	平均值
$2$	90.8	76.7	83.7
$3$	91.7	75.6	83.6
$4$	97.6	71.4	84.5
$5$	97.6	69.1	83.3
$6$	97.6	65.7	81.6

Fig.3 ICA independent component

Fig.4 Fault detection results using ICA

Fig.5 Samples in independent component space by KICA

Fig.6 Fault detection results using KICA

Fig.7 Samples of principal component space in KPCA

Fig.8 Fault detection results using KPCA

Fig.9 Fault detection results using SLKPCA

Fig.10 Samples in feature data using WSFKICA

Fig.11 Fault detection results using WSFKICA

Fig.12 Contribution charts of monitored variables

Fig.13 Monitoring variable

Table 2 Fault detection rates of each method in simulated case

ICA		KICA		KPCA		SLKPCA		WSFKICA
$I 2$	$S P E$	$I 2$	$S P E$	$Τ 2$	$S P E$	$Τ 2$	$S P E$	$R$	$H$
27.6	10.4	27.8	18.6	6.4	2.8	86	87	97.2	92.2

Table 2 Fault detection rates of each method in simulated case

ICA		KICA		KPCA		SLKPCA		WSFKICA
$I 2$	$S P E$	$I 2$	$S P E$	$Τ 2$	$S P E$	$Τ 2$	$S P E$	$R$	$H$
27.6	10.4	27.8	18.6	6.4	2.8	86	87	97.2	92.2

Fig.14 Layout of TE processes

Table 3 Fault detection rates of each method in TE process

序号	ICA		KICA		KPCA		WSFKICA
序号	$I 2$	$S P E$	$I 2$	$S P E$	$Τ 2$	$S P E$	$R$	$H$
1	95.7	99.3	98	98.7	99.1	99.3	99.7	100
2	87.9	94.9	86.6	94	94.4	94.7	93	99.4
3	11.1	10.7	1.6	1.3	0.6	10.3	21	98.6
4	99.9	99.9	99.9	99.9	99.9	99.9	100	100
5	34.7	19.9	4	0.6	1	23.3	43	99.6
6	—	—	—	—	—	—	—	—
7	99.9	99.9	99.9	99.9	99.9	99.9	100	100
8	86.6	88.6	86.4	85.9	87.6	89.4	90.4	93
9	8.7	18.29	2.4	7.3	4.3	13.9	38.3	88.4
10	88	88.6	80.9	84.7	86.9	89.1	93.3	96.7
11	89.4	98.7	95.1	96.6	98.4	98.4	99.4	100
12	32.4	74.1	33.1	35.9	40.4	77	95.1	100
13	92	94.71	93.6	94.1	94.3	94.9	96.7	100
14	83.9	98.9	97.9	98.7	98.7	98.9	100	100
15	14.6	0.4	0.1	0	0	1.9	54.4	99.4
16	5	6.1	6.3	5.9	1.4	0.9	6.7	5.7
17	81.1	84.9	76.3	80	84.7	84.9	88.7	90.7
18	52.9	62.86	48.1	49.6	51.6	62.1	68.7	70.9
19	93.7	95	88.9	90.6	93.1	95.4	94	98.7
20	85.9	85.1	83.6	85.4	85.3	85.3	87.4	91.4
21	5	6.4	6.4	5.7	1.6	0.9	7.3	12.4
22	17.3	13.1	1.6	0.9	0.4	13.1	23.9	87.4
23	16.7	12.6	11.4	18.7	4.7	3.7	44.9	100
24	79.3	87.7	70.1	83.9	83	88.4	91.4	93.1
25	94.1	93.4	83.7	88.3	87	95	98.1	100
26	62.3	87.1	58.1	77.3	85.6	88.6	88.1	95.1
27	50.1	87	59.7	58.4	71.6	90	96.4	100
28	15.7	13	2.4	0.4	0.6	15	3.7	91.7

Table 3 Fault detection rates of each method in TE process

序号	ICA		KICA		KPCA		WSFKICA
序号	$I 2$	$S P E$	$I 2$	$S P E$	$Τ 2$	$S P E$	$R$	$H$
1	95.7	99.3	98	98.7	99.1	99.3	99.7	100
2	87.9	94.9	86.6	94	94.4	94.7	93	99.4
3	11.1	10.7	1.6	1.3	0.6	10.3	21	98.6
4	99.9	99.9	99.9	99.9	99.9	99.9	100	100
5	34.7	19.9	4	0.6	1	23.3	43	99.6
6	—	—	—	—	—	—	—	—
7	99.9	99.9	99.9	99.9	99.9	99.9	100	100
8	86.6	88.6	86.4	85.9	87.6	89.4	90.4	93
9	8.7	18.29	2.4	7.3	4.3	13.9	38.3	88.4
10	88	88.6	80.9	84.7	86.9	89.1	93.3	96.7
11	89.4	98.7	95.1	96.6	98.4	98.4	99.4	100
12	32.4	74.1	33.1	35.9	40.4	77	95.1	100
13	92	94.71	93.6	94.1	94.3	94.9	96.7	100
14	83.9	98.9	97.9	98.7	98.7	98.9	100	100
15	14.6	0.4	0.1	0	0	1.9	54.4	99.4
16	5	6.1	6.3	5.9	1.4	0.9	6.7	5.7
17	81.1	84.9	76.3	80	84.7	84.9	88.7	90.7
18	52.9	62.86	48.1	49.6	51.6	62.1	68.7	70.9
19	93.7	95	88.9	90.6	93.1	95.4	94	98.7
20	85.9	85.1	83.6	85.4	85.3	85.3	87.4	91.4
21	5	6.4	6.4	5.7	1.6	0.9	7.3	12.4
22	17.3	13.1	1.6	0.9	0.4	13.1	23.9	87.4
23	16.7	12.6	11.4	18.7	4.7	3.7	44.9	100
24	79.3	87.7	70.1	83.9	83	88.4	91.4	93.1
25	94.1	93.4	83.7	88.3	87	95	98.1	100
26	62.3	87.1	58.1	77.3	85.6	88.6	88.1	95.1
27	50.1	87	59.7	58.4	71.6	90	96.4	100
28	15.7	13	2.4	0.4	0.6	15	3.7	91.7

Fig.15 Fault detection results using ICA

Fig.16 Fault detection results using KICA

Fig.17 Fault detection results using KPCA

Fig.18 Fault detection results using WSFKICA

Fig.19 Contribution chart of Fault 27

Fig.20 Reactor temperature and reactor cooling water flow

Fig.21 Contribution chart of Fault 10

Fig.22 Stripper temperature and stripper liquid product flow

References 32

1	Kresta J V, MacGregor J F, Marlin T E. Multivariate statistical monitoring of process operating performance[J]. The Canadian Journal of Chemical Engineering, 1991, 69(1): 35-47.
2	刘强, 卓洁, 郎自强, 等. 数据驱动的工业过程运行监控与自优化研究展望[J]. 自动化学报, 2018, 44(11): 1944-1956.
	Liu Q, Zhuo J, Lang Z Q, et al. Perspectives on data-driven operation monitoring and self-optimization of industrial processes[J]. Acta Automatica Sinica, 2018, 44(11): 1944-1956.
3	孔祥玉, 曹泽豪, 杜柏阳, 等. 基于偏最小二乘的质量相关多模态故障检测技术[J]. 控制与决策, 2019, 34(12): 2547-2557.
	Kong X Y, Cao Z H, Du B Y, et al. Quality-related multimodal fault detection technique based on partial least squares[J]. Control and Decision, 2019, 34(12): 2547-2557.
4	Yang Y W, Ma Y X, Song B, et al. An aligned mixture probabilistic principal component analysis for fault detection of multimode chemical processes[J]. Chinese Journal of Chemical Engineering, 2015, 23(8): 1357-1363.
5	冯立伟, 张成, 李元, 等. 基于双近邻标准化和PCA的多阶段过程故障检测[J]. 化工学报, 2018, 69(7): 3159-3166.
	Feng L W, Zhang C, Li Y, et al. DLNS-PCA-based fault detection for multimode batch process[J]. CIESC Journal, 2018, 69(7): 3159-3166.
6	Chen M C, Hsu C C, Malhotra B, et al. An efficient ICA-DW-SVDD fault detection and diagnosis method for non-Gaussian processes[J]. International Journal of Production Research, 2016, 54(17): 5208-5218.
7	Yu J, Yoo J, Jang J, et al. A novel hybrid of auto-associative kernel regression and dynamic independent component analysis for fault detection in nonlinear multimode processes[J]. Journal of Process Control, 2018, 68: 129-144.
8	Teimoortashloo M, Sedigh A K. A modified independent component analysis-based fault detection method in plant-wide systems[J]. Control and Cybernetics, 2015, 44(2): 287-310.
9	Lee J M, Qin S J, Lee I B. Fault detection and diagnosis based on modified independent component analysis[J]. AIChE Journal, 2006, 52(10): 3501-3514.
10	Lee J M, Yoo C, Lee I B. Statistical process monitoring with independent component analysis[J]. Journal of Process Control, 2004, 14(5): 467-485.
11	Choi S W, Lee I B. Nonlinear dynamic process monitoring based on dynamic kernel PCA[J]. Chemical Engineering Science, 2004, 59(24): 5897-5908.
12	赵忠盖, 刘飞. 一种基于核独立元分析的非线性过程监控方法[J]. 系统仿真学报, 2008, 20(20): 5585-5588.
	Zhao Z G, Liu F. Nonlinear process monitoring method based on kernel independent component analysis[J]. Journal of System Simulation, 2008, 20(20): 5585-5588.
13	Lee J M, Qin S J, Lee I B. Fault detection of non-linear processes using kernel independent component analysis[J]. The Canadian Journal of Chemical Engineering, 2007, 85(4): 526-536.
14	Rashid M M, Yu J. Nonlinear and non-Gaussian dynamic batch process monitoring using a new multiway kernel independent component analysis and multidimensional mutual information based dissimilarity approach[J]. Industrial & Engineering Chemistry Research, 2012, 51(33): 10910-10920.
15	周东华, 郭天序, 陈茂银. 基于多次移动平均的微小故障检测方法和装置: 103776480A[P]. 2014-05-07.
	Zhou D H, Guo T X, Chen M Y. Small-fault detection method and device based on multiple moving average: 103776480A[P]. 2014-05-07.
16	邓佳伟, 邓晓刚, 曹玉苹, 等. 基于加权统计局部核主元分析的非线性化工过程微小故障诊断方法[J]. 化工学报, 2019, 70(7): 2594-2605.
	Deng J W, Deng X G, Cao Y P, et al. Incipient fault diagnosis method of nonlinear chemical process based on weighted statistical local KPCA[J]. CIESC Journal, 2019, 70(7): 2594-2605.
17	Zheng K, Luo J F, Zhang Y, et al. Incipient fault detection of rolling bearing using maximum autocorrelation impulse harmonic to noise deconvolution and parameter optimized fast EEMD[J]. ISA Transactions, 2019, 89: 256-271.
18	Wang Y, Tse P W, Tang B P, et al. Kurtogram manifold learning and its application to rolling bearing weak signal detection[J]. Measurement, 2018, 127: 533-545.
19	Wang D, Tsui K L. Dynamic Bayesian wavelet transform: New methodology for extraction of repetitive transients[J]. Mechanical Systems and Signal Processing, 2017, 88: 137-144.
20	Li G Z, Tang G, Luo G G, et al. Underdetermined blind separation of bearing faults in hyperplane space with variational mode decomposition[J]. Mechanical Systems and Signal Processing, 2019, 120: 83-97.
21	Safaeipour H, Forouzanfar M, Ramezani A. Incipient fault detection in nonlinear non-Gaussian noisy environment[J]. Measurement, 2021, 174: 109008.
22	吴强, 孔凡让, 何清波, 等. 基于小波变换和ICA的滚动轴承早期故障诊断[J]. 中国机械工程, 2012, 23(7): 835-840.
	Wu Q, Kong F R, He Q B, et al. Early fault diagnosis of rolling element bearings based on wavelet transform and independent component analysis[J]. China Mechanical Engineering, 2012, 23(7): 835-840.
23	Basseville M. On-board component fault detection and isolation using the statistical local approach[J]. Automatica, 1998, 34(11): 1391-1415.
24	Kruger U, Kumar S, Littler T. Improved principal component monitoring using the local approach[J]. Automatica, 2007, 43(9): 1532-1542.
25	Ge Z Q, Yang C J, Song Z H. Improved kernel PCA-based monitoring approach for nonlinear processes[J]. Chemical Engineering Science, 2009, 64(9): 2245-2255.
26	Chen J, Liao C M. Dynamic process fault monitoring based on neural network and PCA[J]. Journal of Process Control, 2002, 12(2): 277-289.
27	Hyvärinen A, Oja E. Independent component analysis: algorithms and applications[J]. Neural Networks, 2000, 13(4/5): 411-430.
28	Odiowei P P, Cao Y. State-space independent component analysis for nonlinear dynamic process monitoring[J]. Chemometrics and Intelligent Laboratory Systems, 2010, 103(1): 59-65.
29	Shen X, Agrawal S. Kernel density estimation for an anomaly based intrusion detection system[C]// Proceedings of the 2006 International Conference on Machine Learning. Models, Technologies & Applications. Las Vegas, Nevada, USA: DBLP, 2006.
30	邓晓刚, 田学民. 一种基于KPCA的非线性故障诊断方法[J]. 山东大学学报(工学版), 2005, 35(3): 103-106.
	Deng X G, Tian X M. Nonlinear process fault diagnosis method using kernel principal component analysis[J]. Journal of Shandong University (Engineering Science), 2005, 35(3): 103-106.
31	Downs J J, Vogel E F. A plant-wide industrial process control problem[J]. Computers & Chemical Engineering, 1993, 17(3): 245-255.
32	Bathelt A, Ricker N L, Jelali M. Revision of the Tennessee Eastman process model[J]. IFAC-PapersOnLine, 2015, 48(8): 309-314.

[1]	Xin YANG, Wen WANG, Kai XU, Fanhua MA. Simulation analysis of temperature characteristics of the high-pressure hydrogen refueling process [J]. CIESC Journal, 2023, 74(S1): 280-286.
[2]	Jiahao SONG, Wen WANG. Study on coupling operation characteristics of Stirling engine and high temperature heat pipe [J]. CIESC Journal, 2023, 74(S1): 287-294.
[3]	Siyu ZHANG, Yonggao YIN, Pengqi JIA, Wei YE. Study on seasonal thermal energy storage characteristics of double U-shaped buried pipe group [J]. CIESC Journal, 2023, 74(S1): 295-301.
[4]	Fei KANG, Weiguang LYU, Feng JU, Zhi SUN. Research on discharge path and evaluation of spent lithium-ion batteries [J]. CIESC Journal, 2023, 74(9): 3903-3911.
[5]	Yue CAO, Chong YU, Zhi LI, Minglei YANG. Industrial data driven transition state detection with multi-mode switching of a hydrocracking unit [J]. CIESC Journal, 2023, 74(9): 3841-3854.
[6]	Qian MING, Yi GAO, Jian HU, Shengjie LI, Jinjiang WANG. Virtual sensing method for leakage fault of heat exchanger [J]. CIESC Journal, 2023, 74(4): 1836-1846.
[7]	Xiaodan SU, Ganyu ZHU, Huiquan LI, Guangming ZHENG, Ziheng MENG, Fang LI, Yunrui YANG, Benjun XI, Yu CUI. Optimization of wet process phosphoric acid hemihydrate process and crystallization of gypsum [J]. CIESC Journal, 2023, 74(4): 1805-1817.
[8]	Zhongqiu ZHANG, Hongguang LI, Yilin SHI. A multi-task learning approach for complex chemical processes based on manual predictive manipulating strategies [J]. CIESC Journal, 2023, 74(3): 1195-1204.
[9]	Jianghuai ZHANG, Zhong ZHAO. Robust minimum covariance constrained control for C₃ hydrogenation process and application [J]. CIESC Journal, 2023, 74(3): 1216-1227.
[10]	Yuanjing MAO, Zhi YANG, Songping MO, Hao GUO, Ying CHEN, Xianglong LUO, Jianyong CHEN, Yingzong LIANG. Estimation of SAFT-VR Mie equation of state parameters and thermodynamic properties of C₆—C₁₀ alcohols [J]. CIESC Journal, 2023, 74(3): 1033-1041.
[11]	Weiyi SU, Jiahui DING, Chunli LI, Honghai WANG, Yanjun JIANG. Research progress of enzymatic reactive crystallization [J]. CIESC Journal, 2023, 74(2): 617-629.
[12]	Xuejin GAO, Kun CHENG, Huayun HAN, Huihui Gao, Yongsheng QI. Fault diagnosis of chillers using central loss conditional generative adversarial network [J]. CIESC Journal, 2022, 73(9): 3950-3962.
[13]	Jinyu GUO, Zhe WANG, Yuan LI. Fault detection method based on kernel entropy independent component analysis [J]. CIESC Journal, 2022, 73(8): 3647-3658.
[14]	Jing YANG, Zhenkang LIN, Jun TANG, Cheng FAN, Kening SUN. A review of fault characteristics, fault diagnosis and identification for lithium-ion battery systems [J]. CIESC Journal, 2022, 73(8): 3394-3405.
[15]	Le ZHOU, Chengkai SHEN, Chao WU, Beiping HOU, Zhihuan SONG. Deep fusion feature extraction network and its application in chemical process soft sensing [J]. CIESC Journal, 2022, 73(7): 3156-3165.