局部时差约束邻域保持嵌入算法在故障检测中的应用

doi:10.11949/0438-1157.20220210

化工学报 ›› 2022, Vol. 73 ›› Issue (7): 3109-3119.doi: 10.11949/0438-1157.20220210

局部时差约束邻域保持嵌入算法在故障检测中的应用

王琨(),侍洪波(),谭帅,宋冰,陶阳

华东理工大学能源化工过程智能制造教育部重点实验室，上海 200237

收稿日期:2022-02-15 修回日期:2022-04-06 出版日期:2022-07-05 发布日期:2022-08-01
通讯作者: 侍洪波 E-mail:y20190066@mail.ecust.edu.cn;hbshi@ecust.edu.cn
作者简介:王琨（1996—），女，博士研究生，y20190066@mail.ecust.edu.cn
基金资助:
国家重点研发计划项目(2020YFC1522502);国家自然科学基金项目(62073140);上海市青年科技启明星计划项目(21QA1401800);上海市自然科学基金项目(19ZR1473200)

Local time difference constrained neighborhood preserving embedding algorithm for fault detection

Kun WANG(),Hongbo SHI(),Shuai TAN,Bing SONG,Yang TAO

Key Laboratory of Smart Manufacturing in Energy Chemical Process, Ministry of Education, East China University of Science and Technology, Shanghai 200237, China

Received:2022-02-15 Revised:2022-04-06 Published:2022-07-05 Online:2022-08-01
Contact: Hongbo SHI E-mail:y20190066@mail.ecust.edu.cn;hbshi@ecust.edu.cn

摘要/Abstract

摘要：

传统的邻域保持嵌入(neighborhood preserving embedding，NPE)算法通过k近邻(k-nearest neighbors，k-NN)方法选择邻域进行重构来实现降维。但在实际工业过程中采集的样本具有时序相关性，仅仅通过欧氏距离选择近邻样本不能充分反映数据中包含的信息，从而影响检测效果。因此，提出一种局部时差约束邻域保持嵌入(local time difference constrained neighborhood preserving embedding，LTDCNPE)算法，充分考虑样本间的时间和空间关系，从而建立准确的故障检测模型。首先，该算法在固定尺度的时间窗内，根据样本的时序关系和空间特征挑选出邻域。其次，利用样本间的时间差异为邻域样本进行加权，使数据特征保留了高维空间的局部结构。然后，对降维后得到的主元空间和残差空间构建 $T 2$ 和SPE统计量并确定控制限。最后，通过数值例子和Tennessee-Eastman(TE)过程仿真验证LTDCNPE算法的有效性。

关键词: 过程控制, 过程系统, 动态建模, 邻域保持嵌入算法, 邻域选择, 故障检测

Abstract:

The traditional neighborhood preserving embedding (NPE) algorithm uses the k-nearest neighbors (k-NN) method to select neighborhoods for reconstruction to achieve dimensionality reduction. However, the samples collected in the actual industrial process have time series correlation. Selecting the nearest neighbor samples only by Euclidean distance cannot fully reflect the information contained in the data, which will affect the detection performance. Therefore, a local time difference constrained neighborhood preserving embedding (LTDCNPE) algorithm is proposed, which establishes a more accurate fault detection model by fully considering the temporal and spatial relationship between samples. Firstly, the algorithm selects the neighborhoods based on the time and spatial characteristics of the samples within a fixed scale time window. Secondly, the time differences between samples are used to weight the neighborhood samples. In this way, the reconstructed samples retain the local structure of the high-dimensional space. Then, $T 2$ and SPE statistics are calculated for the principal component space and residual space obtained through LTDCNPE. Next, the control limits are determined to detect the process faults. Finally, the performance of LTDCNPE is described by a numerical example and Tennessee Eastman (TE) simulation study.

Key words: process control, process systems, dynamic modeling, neighborhood preserving embedding algorithm, neighborhood selection, fault detection

中图分类号:

TP 277

王琨, 侍洪波, 谭帅, 宋冰, 陶阳. 局部时差约束邻域保持嵌入算法在故障检测中的应用[J]. 化工学报, 2022, 73(7): 3109-3119.

WANG Kun, SHI Hongbo, TAN Shuai, SONG Bing, TAO Yang. Local time difference constrained neighborhood preserving embedding algorithm for fault detection[J]. CIESC Journal, 2022, 73(7): 3109-3119.

图/表 9

图1

图2

表1

过程故障描述"

故障	描述
1	对 $u t$ 引入幅值为2的阶跃故障
2	系数矩阵 $A$ 的第 $2 × 2$ 个元素值由0.264变为1.500，使状态变量 $z t$ 之间的动态关系发生变化

表1

表2

数值例子的漏报率"

Fault	MAR/%
	PCA		NPE		DNPE		LTDCNPE
	$T 2$	SPE	$T 2$	SPE	$T 2$	SPE	$T 2$	SPE
1	58.67	1.33	62.00	2.00	1.00	1.32	0.33	2.00
2	1.67	1.67	1.67	1.67	1.66	1.66	1.39	1.67

表2

图3

图4

表3

TE过程17种故障的漏报率和误报率"

Fault	MAR(FAR)/%
	PCA		NPE		DNPE		LTDCNPE
	$T 2$	SPE	$T 2$	SPE	$T 2$	SPE	$T 2$	SPE
1	0.88(0)	0.13(0.63)	0.88(0)	0.75(0)	0.13(0)	0.50(0)	0.25(1.25)	0.75(0)
2	1.63(1.25)	4(1.25)	1.63(1.25)	1.75(0)	1.25(0)	1.75(0)	1.50(0)	1.75(0)
5	75.88(0.63)	75.88(3.13)	76.25(0.63)	75.38(0.63)	0(1.25)	76.32(0.63)	0(0)	77.25(0.63)
6	0.88(0)	0(1.88)	0.75(0.63)	0(0)	0(1.88)	0(0.63)	0(0)	0(0)
7	0(0)	0(2.50)	0(0)	0(0)	0(1.25)	0(1.25)	0(0.63)	0(0)
8	3.13(0)	13.88(0.63)	3.25(0)	2.50(0)	2.26(0)	2.51(0)	2.25(0)	2.50(0)
10	70.38(0)	70.88(1.25)	70.63(0)	60.63(0)	46.49(0.63)	61.40(0)	12(1.25)	61.13(0)
11	59.38(0.63)	23.88(3.13)	59.25(0.63)	45.50(0.63)	57.39(0.63)	42.61(0)	38.13(0.63)	45.50(0.63)
12	1.63(0)	9.25(3.13)	1.63(0.63)	1.63(0)	0.38(0)	1.00(0)	0.13(1.88)	1.63(0)
13	6.38(0.63)	4.75(1.25)	6.25(0)	5.75(0)	5.51(0)	5.64(0)	4.75(0.63)	5.75(0)
14	0.75(0)	0(1.25)	1.25(0.63)	0.13(0)	0(0.63)	0(0.63)	0(0.63)	0.13(0)
16	86.50(3.75)	67.75(2.50)	84.88(3.13)	78.75(5.63)	55.26(1.88)	81.20(1.88)	8.88(7.50)	79.25(5.63)
17	23.75(1.25)	4.13(2.50)	24.50(1.88)	14.13(0)	14.29(0)	14.29(0)	9.13(0)	14.13(0)
18	10.75(0)	9.75(2.50)	10.63(0)	10.75(0)	10.78(0.63)	10.65(0)	9.63(0.63)	10.75(0)
19	89.00(0)	82.25(0.63)	88.38(0)	98.13(0)	71.43(0)	100(0)	22.00(0.63)	98.13(0)
20	68.25(0)	48.38(4.38)	65.13(0)	57.88(0)	50.50(0)	58.90(0)	11.00(0)	58.38(0)
21	60.75(0)	51.13(5.00)	60.50(0)	61.75(0)	51.13(0.63)	62.91(0)	42.00(3.13)	61.75(0)

表3

图5

图6

参考文献 37

1	Liu J X, Chen J H, Wang D. Linear and exponential fault-assistant feature extraction methods for process monitoring[J]. Control Engineering Practice, 2021, 109(3): 104732.
2	彭开香, 张传放, 马亮, 等. 面向系统层级的复杂工业过程全息故障诊断[J]. 化工学报, 2019, 70(2): 590-598.
	Peng K X, Zhang C F, Ma L, et al. System-levels-based holographic fault diagnosis for complex industrial processes[J]. CIESC Journal, 2019, 70(2): 590-598.
3	Yao L, Ge Z Q. Refining data-driven soft sensor modeling framework with variable time reconstruction[J]. Journal of Process Control, 2020, 87: 91-107.
4	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 T, Tong C D, Shi X H. Variable weighted principal component analysis algorithm and its application in fault detection[J]. CIESC Journal, 2017, 68(8): 3177-3182.
5	Qin S J, Chiang L H. Advances and opportunities in machine learning for process data analytics[J]. Computers & Chemical Engineering, 2019, 126: 465-473.
6	Chang Y Q, Ma R X, Zhao L P, et al. Online operating performance evaluation for the plant-wide industrial process based on a three-level and multi-block method[J]. The Canadian Journal of Chemical Engineering, 2019, 97(S1): 1371-1385.
7	Song B, Shi H B, Tan S, et al. Multisubspace orthogonal canonical correlation analysis for quality-related plant-wide process monitoring[J]. IEEE Transactions on Industrial Informatics, 2021, 17(9): 6368-6378.
8	Tao Y, Shi H B, Song B, et al. A novel dynamic weight principal component analysis method and hierarchical monitoring strategy for process fault detection and diagnosis[J]. IEEE Transactions on Industrial Electronics, 2020, 67(9): 7994-8004.
9	Zhou P, Zhang R Y, Liang M Y, et al. Fault identification for quality monitoring of molten iron in blast furnace ironmaking based on KPLS with improved contribution rate[J]. Control Engineering Practice, 2020, 97: 104354.
10	Li Z C, Yan X F. Fault-relevant optimal ensemble ICA model for non-Gaussian process monitoring[J]. IEEE Transactions on Control Systems Technology, 2019, 28(6): 2581-2590.
11	常玉清, 王姝, 王福利, 等. 基于多PCA模型的过程监测方法[J]. 仪器仪表学报, 2014, 35(4): 901-908.
	Chang Y Q, Wang S, Wang F L, et al. Process monitoring method based on multiple PCA models[J]. Chinese Journal of Scientific Instrument, 2014, 35(4): 901-908.
12	Talmon R, Mallat S, Zaveri H, et al. Manifold learning for latent variable inference in dynamical systems[J]. IEEE Transactions on Signal Processing, 2015, 63(15): 3843-3856.
13	Belkin M, Niyogi P. Laplacian eigenmaps and spectral techniques for embedding and clustering[M]//Advances in Neural Information Processing Systems. Massachusetts, USA: The MIT Press, 2002.
14	Roweis S T, Saul L K. Nonlinear dimensionality reduction by locally linear embedding[J]. Science, 2000, 290(5500): 2323-2326.
15	Tenenbaum J B, Silva V D, Langford J C. A global geometric framework for nonlinear dimensionality reduction[J]. Science, 2000, 290(5500): 2319-2323.
16	He X F, Niyogi P. Locality preserving projections[J]. Advances in Neural Information Processing Systems, 2004, 16(16): 153-160.
17	He X F, Cai D, Yan S C, et al. Neighborhood preserving embedding[C]//Tenth IEEE International Conference on Computer Vision (ICCV'05) Volume 1. Beijing: IEEE, 2005: 1208-1213.
18	Miao A M, Ge Z Q, Song Z H, et al. Nonlocal structure constrained neighborhood preserving embedding model and its application for fault detection[J]. Chemometrics and Intelligent Laboratory Systems, 2015, 142: 184-196.
19	Huang P F, Tao Y, Song B, et al. Tensor sequence component analysis for fault detection in dynamic process[J]. The Canadian Journal of Chemical Engineering, 2020, 98(1): 225-236.
20	Song B, Tan S, Shi H B. Process monitoring via enhanced neighborhood preserving embedding[J]. Control Engineering Practice, 2016, 50: 48-56.
21	Tong C D, Lan T, Shi X H, et al. Statistical process monitoring based on nonlocal and multiple neighborhoods preserving embedding model[J]. Journal of Process Control, 2018, 65: 34-40.
22	Ge Z Q, Chen X R. Dynamic probabilistic latent variable model for process data modeling and regression application[J]. IEEE Transactions on Control Systems Technology, 2019, 27(1): 323-331.
23	Ku W F, Storer R H, Georgakis C. Disturbance detection and isolation by dynamic principal component analysis[J]. Chemometrics and Intelligent Laboratory Systems, 1995, 30(1): 179-196.
24	Li Y, Bi Y, Sun J, et al. Multi-class evaluation using dynamic neighborhood preserving embedding method[J]. Journal of Computational Information Systems, 2015, 11(19): 7001-7006.
25	赵小强, 牟淼. 基于GSFA-GNPE的动态-静态联合指标间歇过程监控[J]. 上海交通大学学报, 2021, 55(11): 1417-1428.
	Zhao X Q, Mou M. Batch process monitoring with dynamic-static joint indicator based on GSFA-GNPE[J]. Journal of Shanghai Jiao Tong University, 2021, 55(11): 1417-1428.
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	Odiowei P P, Cao Y. Nonlinear dynamic process monitoring using canonical variate analysis and kernel density estimations[J]. Computer Aided Chemical Engineering, 2009, 27: 1557-1562.
28	Chen Q, Wynne R J, Goulding P, et al. The application of principal component analysis and kernel density estimation to enhance process monitoring[J]. Control Engineering Practice, 2000, 8(5): 531-543.
29	Liang J. Multivariate statistical process monitoring using kernel density estimation[J]. Developments in Chemical Engineering and Mineral Processing, 2008, 13(1/2): 185-192.
30	Song B, Ma Y X, Shi H B. Multimode process monitoring using improved dynamic neighborhood preserving embedding[J]. Chemometrics and Intelligent Laboratory Systems, 2014, 135:17-30.
31	Zhao X Q, Wang T. Tensor dynamic neighborhood preserving embedding algorithm for fault diagnosis of batch process[J]. Chemometrics and Intelligent Laboratory Systems, 2017, 162: 94-103.
32	Bathelt A, Ricker N L, Jelali M. Revision of the Tennessee Eastman process model[J]. IFAC-PapersOnLine, 2015, 48(8): 309-314.
33	Rong M Y, Shi H B, Tan S. Large-scale supervised process monitoring based on distributed modified principal component regression[J]. Industrial & Engineering Chemistry Research, 2019, 58(39): 18223-18240.
34	Lyman P R, Georgakis C. Plant-wide control of the Tennessee Eastman problem[J]. Computers & Chemical Engineering, 1995, 19(3): 321-331.
35	Yu J B. Local and global principal component analysis for process monitoring[J]. Journal of Process Control, 2012, 22(7): 1358-1373.
36	Saul L, Roweis S. Think globally, fit locally: unsupervised learning of low dimensional manifold[J]. Journal of Machine Learning Research, 2003, 4: 119-155.
37	蒋浩天, E.L. 拉塞尔, R.D. 布拉茨, 等. 工业系统的故障检测与诊断[M]. 段建民, 译. 北京: 机械工业出版社, 2003.
	Chiang L H, Russell E L, Braatz R D, et al. Fault Detection and Diagnosis in Industrial Systems[M]. Duan J M, trans. Beijing: China Machine Press, 2003.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

[1]	杨岭, 崔国民, 周志强, 肖媛. 精细搜索策略应用于质量交换网络综合[J]. 化工学报, 2022, 73(7): 3145-3155.
[2]	赵涛岩, 曹江涛, 李平, 冯琳, 商瑀. 区间二型模糊免疫PID在环己烷无催化氧化温度控制系统中的应用[J]. 化工学报, 2022, 73(7): 3166-3173.
[3]	周乐, 沈程凯, 吴超, 侯北平, 宋执环. 深度融合特征提取网络及其在化工过程软测量中的应用[J]. 化工学报, 2022, 73(7): 3156-3165.
[4]	孟文亮, 李贵贤, 周怀荣, 李婧玮, 王健, 王可, 范学英, 王东亮. 绿氢重构的粉煤气化煤制甲醇近零碳排放工艺研究[J]. 化工学报, 2022, 73(4): 1714-1723.
[5]	张淑君, 王诗慧, 张欣, 吉旭, 戴一阳, 党亚固, 周利. 集成轻烃回收单元代理模型的氢气网络多目标优化[J]. 化工学报, 2022, 73(4): 1658-1672.
[6]	王建松, 许锋, 罗雄麟. 化工过程多回路PID控制系统模式切换参数自整定[J]. 化工学报, 2022, 73(4): 1647-1657.
[7]	张欣, 周利, 王诗慧, 吉旭, 毕可鑫. 考虑原油性质波动的炼厂氢气网络集成优化[J]. 化工学报, 2022, 73(4): 1631-1646.
[8]	高欢, 丁国良, 周发贤, 庄大伟. R410A制冷剂在润滑油中的动态析出特性的研究[J]. 化工学报, 2022, 73(3): 1054-1062.
[9]	张建飞, 林嘉奖, 罗雄麟, 许锋. 重油催化裂化装置产品分布调控与优化模拟分析[J]. 化工学报, 2022, 73(3): 1232-1245.
[10]	曹森山, 许锋, 罗雄麟. 基于稳定性的循环物流系统流程模拟——以催化裂化反应-再生系统为例[J]. 化工学报, 2022, 73(3): 1256-1269.
[11]	张兴硕, 罗雄麟, 许锋. 催化裂化装置反再系统动态模拟精细化与控制系统“工艺优先”配对设计[J]. 化工学报, 2022, 73(2): 747-758.
[12]	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.
[13]	徐健玮, 梁颖宗, 罗向龙, 陈健勇, 杨智, 陈颖. 液化天然气深冷-膜蒸馏海水淡化系统集成与分析[J]. 化工学报, 2021, 72(S1): 437-444.
[14]	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.
[15]	卢道铭, 唐钊艇, 范怡平, 卢春喜. 大差异颗粒分级再生设备的性能研究[J]. 化工学报, 2021, 72(8): 4184-4195.

局部时差约束邻域保持嵌入算法在故障检测中的应用

Local time difference constrained neighborhood preserving embedding algorithm for fault detection

RichHTML

PDF (PC)

摘要/Abstract

引用本文

使用本文

图/表 9

参考文献 37

相关文章 15

Metrics

本文评价

推荐阅读 0