Process monitoring method based on LSTM adaptive dynamic reconstruction principal component analysis

GAO Zhenpeng; ZHANG Haonan; ZHAO Lijie

doi:10.11949/0438-1157.20260915

您当前的位置：

首页 >

文章列表页 >

Process monitoring method based on LSTM adaptive dynamic reconstruction principal component analysis

更新时间：2026-08-19

- Process monitoring method based on LSTM adaptive dynamic reconstruction principal component analysis
- CIESC Journal (2026)
- 作者机构：
  
  沈阳化工大学信息工程学院，辽宁沈阳 110142
- 作者简介：
- 基金信息：
- DOI：10.11949/0438-1157.20260915
  CLC： TQ 028.8
- Received：01 July 2026，
  
  Revised：2026-08-17，
  
  Accepted：19 August 2026，
- 稿件说明：
移动端阅览
GAO Zhenpeng, ZHANG Haonan, ZHAO Lijie. Process monitoring method based on LSTM adaptive dynamic reconstruction principal component analysis[J/OL]. CIESC Journal, 2026.
DOI：

GAO Zhenpeng, ZHANG Haonan, ZHAO Lijie. Process monitoring method based on LSTM adaptive dynamic reconstruction principal component analysis[J/OL]. CIESC Journal, 2026. DOI： 10.11949/0438-1157.20260915.

摘要

针对间歇过程变量时序相关性强、运行状态时变以及早期故障信号弱等问题，传统动态重构主成分分析（DRPCA）采用静态重构矩阵，难以跟踪不同运行阶段的动态特性变化；同时，线性PCA对增量矩阵的表征能力有限，导致微小故障和过程监测灵敏度不足，易出现检出延迟和漏报率偏高等现象。为此，本文提出一种基于长短期记忆网络（LSTM）的自适应动态重构主成分分析方法（LADRPCA），用于过程监测。该方法首先利用LSTM对滑动窗口时序数据建模，获取过程变量的长短期时间相关性；随后结合Cayley变换在线生成满足正交约束的时变重构矩阵，使模型能够随过程状态变化自适应更新；最后在PCA统计量监测框架的基础上，引入自编码器对增量矩阵进行特征提取，并利用重构误差增强对早期微弱故障的监测能力。在三水箱过程和青霉素发酵过程上的实验结果表明，LADRPCA能够缩短故障检测延迟，提高故障检测率，并在较低误报率下保持对早期微弱故障的较好灵敏度与稳定性，具有一定的过程监测应用潜力。

Abstract

Batch processes are characterized by strong temporal correlations among process variables

time-varying operating states

and weak incipient fault signals. Traditional dynamic reconstruction principal component analysis (DRPCA) employs a static reconstruction matrix

making it difficult to track changes in dynamic characteristics across different operating phases. Moreover

linear PCA has limited capability to represent incremental matrices

resulting in insufficient sensitivity to minor and incipient faults

as well as delayed detection and missed alarms. To address these problems

an LSTM-based adaptive dynamic reconstruction principal component analysis method

termed LADRPCA

is proposed for process monitoring. First

an LSTM network is employed to model sliding-window time-series data and capture the long- and short-term temporal correlations among process variables. Then

the Cayley transform is used to generate time-varying reconstruction matrices satisfying orthogonality constraints

enabling the model to adaptively track changes in process states. Finally

an autoencoder is incorporated into the PCA statistical monitoring framework to extract features from incremental matrices

and its reconstruction error is used to enhance the sensitivity to incipient weak faults. Experimental results obtained from a three-tank process and a penicillin fermentation process demonstrate that LADRPCA can reduce fault detection delays

improve fault detection rates

and maintain good sensitivity and stability to incipient faults at a low false alarm rate

indicating its potential for process monitoring applications.

关键词

Keywords

references

周东华 , 史建涛 , 何潇 . 动态系统间歇故障诊断技术综述 [J ] . 自动化学报 , 2014 , 40 ( 2 ): 161 - 171 .

Zhou D H , Shi J T , He X . Review of intermittent fault diagnosis techniques for dynamic systems [J ] . Acta Automatica Sinica , 2014 , 40 ( 2 ): 161 - 171 .

Van den Kerkhof P , Gins G , Vanlaer J , et al . Dynamic model-based fault diagnosis for (bio)chemical batch processes [J ] . Computers & Chemical Engineering , 2012 , 40 : 12 - 21 .

Marais H , Van S G , Uren K R . The merits of exergy-based fault detection in petrochemical processes [J ] . Journal of Process Control , 2019 , 74 : 110 - 119 .

Baklouti I , Mansouri M , Ben Hamida A , et al . Monitoring of wastewater treatment plants using improved univariate statistical technique [J ] . Process Safety and Environmental Protection , 2018 , 116 : 287 - 300 .

Strydom J J , Miskin J J , McCoy J T , et al . Fault diagnosis and economic performance evaluation for a simulated base metal leaching operation [J ] . Minerals Engineering , 2018 , 123 : 128 - 143 .

Fan S S , Hsu C Y , Tsai D M , et al . Data-driven approach for fault detection and diagnostic in semiconductor manufacturing [J ] . IEEE Transactions on Automation Science and Engineering , 2020 , 17 ( 4 ): 1925 - 1936 .

Tulsyan A , Garvin C , Ündey C . Advances in industrial biopharmaceutical batch process monitoring: Machine-learning methods for small data problems [J ] . Biotechnology and Bioengineering , 2018 , 115 ( 8 ): 1915 - 1924 .

Kong X Y , Yang Z Y , Liu Y M , et al . Overview of industrial process monitoring methods based on independent component analysis and its extended model [J ] . Control and Decision , 2022 , 37 ( 4 ): 799 - 814 .

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 .

Wang L M , He X , Zhou D H . Average dwell time-based optimal iterative learning control for multi-phase batch processes [J ] . Journal of Process Control , 2016 , 40 : 1 - 12 .

Zhang H L , Qi Y S , Wang L , et al . Fault detection and diagnosis of chemical process using enhanced KECA [J ] . Chemometrics and Intelligent Laboratory Systems , 2017 , 161 : 61 - 69 .

Ben Khediri I , Limam M , Weihs C . Variable window adaptive kernel principal component analysis for nonlinear nonstationary process monitoring [J ] . Computers & Industrial Engineering , 2011 , 61 ( 3 ): 437 - 446 .

Jackson J E . Quality control methods for several related variables [J ] . Technometrics , 1959 , 1 ( 4 ): 359 - 377 .

Jackson J E , Mudholkar G S . Control procedures for residuals associated with principal component analysis [J ] . Technometrics , 1979 , 21 ( 3 ): 341 - 349 .

Ji C , Sun W . A review on data-driven process monitoring methods: characterization and mining of industrial data [J ] . Processes , 2022 , 10 ( 2 ): 335 .

Melo A , Câmara M M , Pinto J C . Data-driven process monitoring and fault diagnosis: a comprehensive survey [J ] . Processes , 2024 , 12 ( 2 ): 251 .

Jolliffe I T , Cadima J . Principal component analysis: a review and recent developments [J ] . Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences , 2016 , 374 ( 2065 ): 20150202 .

Sheriff M Z , Botre C , Mansouri M , et al . Process monitoring using data-based fault detection techniques: comparative studies [M ] . Fault Diagnosis and Detection . 2017 , 32 : 137 - 144 .

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 .

Wold S , Sjöström M , Eriksson L . PLS-regression: a basic tool of chemometrics [J ] . Chemometrics and Intelligent Laboratory Systems , 2001 , 58 ( 2 ): 109 - 130 .

Apsemidis A , Psarakis S , Moguerza J M . A review of machine learning kernel methods in statistical process monitoring [J ] . Computers & Industrial Engineering , 2020 , 142 : 106376 .

Pani A K . Non-linear process monitoring using kernel principal component analysis: a review of the basic and modified techniques with industrial applications [J ] . Brazilian Journal of Chemical Engineering , 2022 , 39 ( 2 ): 327 - 344 .

Si Y B , Wang Y Q , Zhou D H . Key-performance-indicator-related process monitoring based on improved kernel partial least squares [J ] . IEEE Transactions on Industrial Electronics , 2021 , 68 ( 3 ): 2626 - 2636 .

顾炳斌 , 熊伟丽 . 基于多块信息提取的PCA故障诊断方法 [J ] . 化工学报 , 2019 , 70 ( 2 ): 736 - 749 .

Gu B B , Xiong W L . Fault diagnosis method for PCA based on multi-block information extraction [J ] . CIESC Journal , 2019 , 70 ( 2 ): 736 – 749 .

Chen X L , Wang J , Ding S X . Complex system monitoring based on distributed least squares method [J ] . IEEE Transactions on Automation Science and Engineering , 2021 , 18 ( 4 ): 1892 - 1900 .

Yu W K , Zhao C H , Huang B . MoniNet with concurrent analytics of temporal and spatial information for fault detection in industrial processes [J ] . IEEE Transactions on Cybernetics , 2022 , 52 ( 8 ): 8340 - 8351 .

Yu W K , Wu M , Huang B , et al . A generalized probabilistic monitoring model with both random and sequential data [J ] . Automatica , 2022 , 144 : 110468 .

Cohen A , Atoui M A . On wavelet-based statistical process monitoring [J ] . Transactions of the Institute of Measurement and Control , 2022 , 44 ( 3 ): 525 - 538 .

衷路生 , 何东 , 龚锦红 , 等 . 基于分布式ICA-PCA模型的工业过程故障监测 [J ] . 化工学报 , 2015 , 66 ( 11 ): 4546 - 4554 .

Zhong L S , He D , Gong J H , et al . Fault monitoring of industrial process based on distributed ICA-PCA model [J ] . CIESC Journal , 2015 , 66 ( 11 ): 4546 - 4554 .

Harrou F , Madakyaru M , Sun Y . Improved nonlinear fault detection strategy based on the Hellinger distance metric: Plug flow reactor monitoring [J ] . Energy and Buildings , 2017 , 143 : 149 - 161 .

高学金 , 黄梦丹 , 齐咏生 , 等 . PDPSO优化多阶段AR-PCA间歇过程监测方法 [J ] . 化工学报 , 2018 , 69 ( 9 ): 3914 - 3923 .

Gao X J , Huang M D , Qi Y S , et al . Batch process monitoring using multiphase AR-PCA optimized with PDPSO [J ] . CIESC Journal , 2018 , 69 ( 9 ): 3914 - 3923 .

Dong Y N , Qin S J . A novel dynamic PCA algorithm for dynamic data modeling and process monitoring [J ] . Journal of Process Control , 2018 , 67 : 1 - 11 .

Liu F , Wang P L , Cai Z D , et al . Batch process fault diagnosis based on the combination of deep belief network and long short-term memory network [C ] // 2019 CAA Symposium on Fault Detection, Supervision and Safety for Technical Processes (SAFEPROCESS) . July 5 - 7 , 2019 . Xiamen, China . IEEE , 2019: 208 - 214 .

Yu J B , Zheng X Y , Liu J T . Stacked convolutional sparse denoising auto-encoder for identification of defect patterns in semiconductor wafer map [J ] . Computers in Industry , 2019 , 109 : 121 - 133 .

朱春梦 , 李增 , 柳楠 , 等 . 基于自编码器和多尺度符号转移熵的FCC沉降器跑剂故障检测 [J ] . 化工学报 , 2025 , 76 ( 9 ): 4512 - 4523 .

Zhu C M , Li Z , Liu N , et al . Fault detection of catalyst loss in FCC disengager based on autoencoder and multi-scale symbolic transfer entropy [J ] . CIESC Journal , 2025 , 76 ( 9 ): 4512 - 4523 .

高学金 , 刘腾飞 , 徐子东 , 等 . 基于循环自动编码器的间歇过程故障监测 [J ] . 化工学报 , 2020 , 71 ( 7 ): 3172 - 3179 .

Gao X J , Liu T F , Xu Z D , et al . Intermittent process fault monitoring based on recurrent autoencoder [J ] . CIESC Journal , 2020 , 71 ( 7 ): 3172 - 3179 .

张海利 , 王普 , 高学金 , 等 . 基于批次图像化的卷积自编码故障监测方法 [J ] . 控制与决策 , 2021 , 36 ( 6 ): 1361 - 1367 .

Zhang H L , Wang P , Gao X J , et al . Fault detection of batch image-based convolutional autoencoder [J ] . Control and Decision , 2021 , 36 ( 6 ): 1361 - 1367 .

Agarwal P , Aghaee M , Tamer M , et al . A novel unsupervised approach for batch process monitoring using deep learning [J ] . Computers & Chemical Engineering , 2022 , 159 : 107694 .

Wang K , Gopaluni R B , Chen J , et al . Deep learning of complex batch process data and its application on quality prediction [J ] . IEEE Transactions on Industrial Informatics , 2020 , 16 ( 12 ): 7233 - 7242 .

Li H Q , Jia M X , Mao Z Z . Dynamic reconstruction principal component analysis for process monitoring and fault detection in the cold rolling industry [J ] . Journal of Process Control , 2023 , 128 : 103010 .

Birol G , Ündey C , Çinar A . A modular simulation package for fed-batch fermentation: penicillin production [J ] . Computers & Chemical Engineering , 2002 , 26 ( 11 ): 1553 - 1565 .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Intermittent process monitoring based on GSA-LSTM dynamic structure feature extraction

MWPCA blast furnace anomaly monitoring algorithm based on Gaussian mixture model

Fault detection method based on state space-PCANet

Fault detection method based on minimum sufficient statistics pattern analysis

Quality-related fault detection based on weighted mutual information principal component analysis

Related Author

WANG Yalin

PAN Yuqing

LIU Chenliang

ZHU Xiongzhuo

ZHANG Hanwen

YANG Chunjie

DONG Shun

LI Yiguo

Related Institution

School of Automation, Central South University

College of Control Science and Engineering, Zhejiang University

School of Energy and Environment, Southeast University

Key Laboratory of Energy Thermal Conversion and Control of Ministry of Education, Southeast University

Jiangsu Frontier Electric Technology Company Limited

⁰