Fault detection of batch process based on double weight and multiple neighborhoods preserving embedding

被引：0

作者：

Yao H.-J. ^{[1
,2
,3
]}

Zhao X.-Q. ^{[1
,2
,3
]}

Li W. ^{[1
,2
,3
]}

Hui Y.-Y. ^{[1
,2
,3
]}

机构：

[1] College of Electrical and Information Engineering, Lanzhou University of Technology, Lanzhou

[2] Gansu Key Laboratory of Advanced Control for Industrial Processes, Lanzhou University of Technology, Lanzhou

[3] National Experimental Teaching Center of Electrical and Control Engineering, Lanzhou University of Technology, Lanzhou

来源：

Zhao, Xiao-Qiang (xqzhao@lut.cn); Zhao, Xiao-Qiang (xqzhao@lut.cn); Zhao, Xiao-Qiang (xqzhao@lut.cn) | 1600年 / Northeast University卷 / 36期

关键词：

Angle neighbor; Batch process; Dynamic; Fault detection; Local outlier factor; NPE;

D O I：

10.13195/j.kzyjc.2020.0659

中图分类号：

学科分类号：

摘要：

Aiming at the problem of fault detection caused by the dynamic characteristic of batch process data, a double weight multiple neighborhoods preserving embedding (DWMNPE) algorithm is proposed. Firstly, by finding time neighbors for each sample point, the time correlations between samples are reflected. By defining angle neighbors, sample points are reconstructed to represent the similarity in the space by finding time neighbors, angle neighbors and distance neighbors for sample points. Three different manifold features can fully extract the essential structure of original data. Then, considering the minimum error and three kinds of neighbor order information, a new objective function is constructed to further prevent the loss of neighbor order information when the reconstructing weights are calculated using the NPE algorithm. The data dynamic is solved, meanwhile, the essential local structure is achieved. Finally, the LOF statistic of the dimensionality reduction data is constructed to monitor the process and eliminate the bad influence of data non-Gaussian for monitoring effect. The results of a numerical example and the penicillin fermentation process simulation demonstrate that the DWMNPE algorithm is effective for fault detection in dynamic batch process. © 2021, Editorial Office of Control and Decision. All right reserved.

引用

页码：3023 / 3030

页数：7

共 26 条

[1] Zhao C H, Yu W K, Gao F R., Data analytics and condition monitoring methods for nonstationary batch processes - Currrent status and future, Acta Automatica Sinica, 46, 10, pp. 2072-2091, (2020)
[2] Lu N Y, Wang F L, Gao F R, Et al., Statistical modeling and online monitoring for batch processes, Acta Automatica Sinica, 32, 3, pp. 400-410, (2006)
[3] Ge Z Q, Song Z H, Gao F R., Review of recent research on data-based process monitoring, Industrial & Engineering Chemistry Research, 52, 10, pp. 3543-3562, (2013)
[4] Chang P, Wang P, Gao X J., Fault monitoring batch process based on statistics pattern analysis of T-KPLS, CIESC Journal, 66, 1, pp. 265-271, (2015)
[5] Tong C D, Lan T, Shi X H., Fault detection and diagnosis of dynamic processes using weighted dynamic decentralized PCA approach, Chemometrics and Intelligent Laboratory Systems, 161, pp. 34-42, (2017)
[6] Kong X Y, Cao Z H, An Q S, Et al., Review of partial least squares linear models and their nonlinear dynamic expansion models, Control and Decision, 33, 9, pp. 1537-1548, (2018)
[7] Tenenbaum J B, de Silva V, Langford J C., A global geometric framework for nonlinear dimensionality reduction, Science, 290, 5500, pp. 2319-2323, (2000)
[8] Belkin M, Niyogi P., Laplacian eigenmaps and spectral techniques for embedding and clustering, Advances in Neural Information Processing Systems, 14, 6, pp. 585-591, (2002)
[9] Roweis S T., Nonlinear dimensionality reduction by locally linear embedding, Science, 290, 5500, pp. 2323-2326, (2000)
[10] He X F, Cai D, Yan S C, Et al., Neighborhood preserving embedding, Proceeding of the 10th IEEE International Conference on Computer Vision, pp. 1208-1213, (2005)

← 1 2 3 →