Application of adaptive algorithm of online reduced KECA in fault detection

doi:10.11949/0438-1157.20201667

Abstract

Abstract:

In complex large-scale industrial process system, real-time process monitoring, computational optimization and reduction of running memory are the most critical and challenging tasks to achieve final product quality, so an adaptive fault detection algorithm for online reduced kernel entropy component analysis (ORKECA) is proposed. First, the kernel matrix of the training samples is calculated. The representative observation values are selected according to the retained eigenvalues and eigenvectors to construct a reduced set that conforms to the global data information characteristics. The square prediction error (SPE) of monitoring statistical data is calculated, and the control limit is determined by kernel density estimation. For the real-time data collected online, the statistic of the data is calculated and compared with the control limit of the reduced set. Whether the kernel entropy component analysis (KECA) model needs to be updated is analyzed according to the process state. This method can improve the performance of real-time monitoring process data. Finally, a non-linear numerical case and TE process data are used to simulate and numerically analyze the method. The simulation results show that the proposed method is effective and feasible.

Key words: optimization, fault detection, kernel entropy component analysis, adaptive algorithm, kernel density estimation, model, numerical analysis

摘要：

在复杂的大规模工业过程系统中，实时过程监视、优化计算时间和降低运行内存是实现最终产品质量的最关键和最具挑战性的任务，提出一种在线压缩核熵成分分析（online reduced kernel entropy component analysis， ORKECA）的自适应故障检测算法。首先计算训练样本的核矩阵，根据保留的特征值与特征向量选择有代表性的观测值，构造一个符合全局数据信息特征的压缩集，计算监测统计数据的平方预测误差（squared prediction error， SPE），并利用核密度估计确定控制限。对于在线实时采集的数据，计算该数据的统计量并与压缩集的控制限比较，根据过程状态分析核熵成分分析（kernel entropy component analysis， KECA）模型是否需要进行更新，可以有效提高实时监测过程数据的性能。最后，以一个非线性数值案例及TE过程数据对该方法进行仿真数值分析。结果表明，所提的方法具有有效的可行性。

关键词: 优化, 故障检测, 核熵成分分析, 自适应算法, 核密度估计, 模型, 数值分析

CLC Number:

TP 277

Jinyu GUO, Wentao LI, Yuan LI. Application of adaptive algorithm of online reduced KECA in fault detection[J]. CIESC Journal, 2021, 72(8): 4227-4238.

郭金玉, 李文涛, 李元. 在线压缩KECA的自适应算法在故障检测中的应用[J]. 化工学报, 2021, 72(8): 4227-4238.

Figures/Tables 10

References 31

1	Amar M, Gondal I, Wilson C. Vibration spectrum imaging: a novel bearing fault classification approach[J]. IEEE Transactions on Industrial Electronics, 2015, 62(1): 494-502.
2	李晗, 萧德云. 基于数据驱动的故障诊断方法综述[J]. 控制与决策, 2011, 26(1): 1-9, 16.
	Li H, Xiao D Y. Survey on data driven fault diagnosis methods[J]. Control and Decision, 2011, 26(1): 1-9, 16.
3	康德礼, 李秀喜. 核熵成分分析算法及其在化工过程监控中的应用研究[J]. 计算机与应用化学, 2013, 30(8): 855-858.
	Kang D L, Li X X. Chemical process monitoring using kernel entropy component[J]. Computers and Applied Chemistry, 2013, 30(8): 855-858.
4	Qin S J. Survey on data-driven industrial process monitoring and diagnosis[J]. Annual Reviews in Control, 2012, 36(2): 220-234.
5	Liu Y H, Sun R, Jin S. A survey on data-driven process monitoring and diagnostic methods for variation reduction in multi-station assembly systems[J]. Assembly Automation, 2019, 39(4): 727-739.
6	Yin S, Ding S X, Xie X C, et al. A review on basic data-driven approaches for industrial process monitoring[J]. IEEE Transactions on Industrial Electronics, 2014, 61(11): 6418-6428.
7	尚宇炜, 马钊, 彭晨阳, 等. 内嵌专业知识和经验的机器学习方法探索(一): 引导学习的提出与理论基础[J]. 中国电机工程学报, 2017, 37(19): 5560-5571, 5833.
	Shang Y W, Ma Z, Peng C Y, et al. Study of a novel machine learning method embedding expertise(Ⅰ): Proposals and fundamentals of guiding learning[J]. Proceedings of the CSEE, 2017, 37(19): 5560-5571, 5833.
8	Xiong L, Liang J, Qian J X. Multivariate statistical process monitoring of an industrial polypropylene catalyzer reactor with component analysis and kernel density estimation[J]. Chinese Journal of Chemical Engineering, 2007, 15(4): 524-532.
9	Westerhuis J A, Gurden S P, Smilde A K. Generalized contribution plots in multivariate statistical process monitoring[J]. Chemometrics and Intelligent Laboratory Systems, 2000, 51(1): 95-114.
10	Wang J, He Q P. Multivariate statistical process monitoring based on statistics pattern analysis[J]. Industrial & Engineering Chemistry Research, 2010, 49(17): 7858-7869.
11	Zhu J L, Ge Z Q, Song Z H. Distributed parallel PCA for modeling and monitoring of large-scale plant-wide processes with big data[J]. IEEE Transactions on Industrial Informatics, 2017, 13(4): 1877-1885.
12	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.
13	Russell E L, Chiang L H, Braatz R D. Fault detection in industrial processes using canonical variate analysis and dynamic principal component analysis[J]. Chemometrics and Intelligent Laboratory Systems, 2000, 51(1): 81-93.
14	Lee C, Choi S W, Lee I B. Variable reconstruction and sensor fault identification using canonical variate analysis[J]. Journal of Process Control, 2006, 16(7): 747-761.
15	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.
16	刘舒锐, 彭慧, 李帅, 等. 基于IJB-PCA-ICA算法的故障检测[J]. 化工学报, 2018, 69(12): 5146-5154.
	Liu S R, Peng H, Li S, et al. Fault detection based on IJB-PCA-ICA[J]. CIESC Journal, 2018, 69(12): 5146-5154.
17	Cho J H, Lee J M, Wook Choi S, et al. Fault identification for process monitoring using kernel principal component analysis[J]. Chemical Engineering Science, 2005, 60(1): 279-288.
18	Zhao C H, Gao F R, Wang F L. Nonlinear batch process monitoring using phase-based kernel-independent component analysis-principal component analysis (KICA-PCA)[J]. Industrial & Engineering Chemistry Research, 2009, 48(20): 9163-9174.
19	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.
20	Zhang Y W, Ma C. Fault diagnosis of nonlinear processes using multiscale KPCA and multiscale KPLS[J]. Chemical Engineering Science, 2011, 66(1): 64-72.
21	Chen J Y, Yu J. Independent component analysis mixture model based dissimilarity method for performance monitoring of non-Gaussian dynamic processes with shifting operating conditions[J]. Industrial & Engineering Chemistry Research, 2014, 53(13): 5055-5066.
22	Jenssen R. Kernel entropy component analysis[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2010, 32(5): 847-860.
23	Jiang Q C, Yan X F, Lv Z, et al. Fault detection in nonlinear chemical processes based on kernel entropy component analysis and angular structure[J]. Korean Journal of Chemical Engineering, 2013, 30(6): 1181-1186.
24	Gómez-Chova L, Jenssen R, Camps-Valls G. Kernel entropy component analysis for remote sensing image clustering[J]. IEEE Geoscience and Remote Sensing Letters, 2012, 9(2): 312-316.
25	Jaffel I, Taouali O, Harkat M F, et al. Kernel principal component analysis with reduced complexity for nonlinear dynamic process monitoring[J]. The International Journal of Advanced Manufacturing Technology, 2017, 88(9/10/11/12): 3265-3279.
26	Binol H. Ensemble learning based multiple kernel principal component analysis for dimensionality reduction and classification of hyperspectral imagery[J]. Mathematical Problems in Engineering, 2018, 2018: 1-14.
27	Downs J J, Vogel E F. A plant-wide industrial process control problem[J]. Computers & Chemical Engineering, 1993, 17(3): 245-255.
28	文莹, 闫雅慧. 基于自适应核主元分析的航空发动机异常监测[J]. 兵工自动化, 2016, 35(8): 1-4, 8.
	Wen Y, Yan Y H. Aero engine anomaly monitoring based on self-adaptive kernel principal component analysis[J]. Ordnance Industry Automation, 2016, 35(8): 1-4, 8.
29	Lee J M, Yoo C, Choi S W, et al. Nonlinear process monitoring using kernel principal component analysis[J]. Chemical Engineering Science, 2004, 59(1): 223-234.
30	Tracy N D, Young J C, Mason R L. Multivariate control charts for individual observations[J]. Journal of Quality Technology, 1992, 24(2): 88-95.
31	Sullivan J H, Woodall W H. A comparison of multivariate control charts for individual observations[J]. Journal of Quality Technology, 1996, 28(4): 398-408.

故障类型	ET/s
故障类型	KECA	RKECA	ORKECA
故障1	0.0374	0.0101	0.0240
故障2	0.0311	0.0173	0.0301

故障类型	ET/s
故障类型	KECA	RKECA	ORKECA
故障1	0.0374	0.0101	0.0240
故障2	0.0311	0.0173	0.0301

故障类型	ET/s
故障类型	KECA	RKECA	ORKECA
故障2	0.0214	0.0109	0.0197
故障17	0.0185	0.0093	0.0133

故障类型	ET/s
故障类型	KECA	RKECA	ORKECA
故障2	0.0214	0.0109	0.0197
故障17	0.0185	0.0093	0.0133

故障类型	KPCA		MWKPCA		KECA		RKECA		ORKECA
故障类型	FAR/%	FDR/%	FAR/%	FDR/%	FAR/%	FDR/%	FAR/%	FDR/%	FAR/%	FDR/%
1	5.63	99.50	1.25	99.75	6.25	99.79	12.50	99.79	0	99.88
2	3.75	98.63	5.00	98.12	4.37	98.85	12.63	98.96	2.25	99.38
4	8.75	88.00	5.00	92.88	6.25	93.00	9.85	94.13	5.50	96.25
5	8.75	30.88	1.25	63.25	6.26	39.79	7.87	57.61	1.25	72.55
6	2.50	99.25	6.25	99.75	3.13	99.65	9.50	98.16	1.25	99.75
7	1.88	100.00	1.25	100	0.63	100.00	5.75	99.89	0.30	100.00
8	8.13	97.50	3.13	97.13	8.13	98.00	13.50	99.58	2.50	100.00
10	3.13	34.38	4.26	57.37	2.50	43.50	14.63	76.88	3.38	74.88
11	5.63	65.38	6.25	74.00	4.33	72.6	15.50	84.30	5.00	90.50
12	4.38	96.88	1.88	98.13	3.75	96.35	10.25	98.85	4.63	99.63
13	1.88	95.88	9.18	99.88	1.25	96.56	8.75	98.02	1.88	100.00
14	3.13	100	10.50	100.00	5.00	100.00	7.13	100.00	2.50	100.00
16	5.00	24.25	12.25	40.25	5.63	34.79	8.60	36.50	6.50	52.63
17	5.63	86.88	0.63	87.50	3.75	89.06	8.75	91.98	1.87	97.00
18	4.35	90.75	2.50	93.38	5.63	92.35	11.25	93.54	3.25	95.00
19	2.50	26.00	5.25	42.13	3.75	33.75	13.30	40.10	7.00	48.50
20	1.88	45.25	1.88	40.88	1.25	54.95	6.13	60.97	0.50	59.75
21	6.88	46.88	12.13	66.75	5.63	57.00	11.88	59.97	8.25	62.25