检索结果-南通市图书馆

Cross-Domain Industrial process monitoring Based on Multiple Structured Latent Transfer Dictionary Pair Learning

第35届中国过程控制会议

作者： Hongyu Tian Zhengheng Ding Hongbo Shi School of Information Science and Engineering East China University of Science and Technology

With the advancement of modern industry, the coupling and information interaction among process variables have become increasingly complex. Current key-performance-indicators-related(KPIs) process monitoring methods typically assume that the training data is sufficient and ignore the scarcity of labeled data. Additionally, not all variables are correlated with KPIs, making it inappropriate to use all variables to construct statistic for fault detection and to directly locate faults using contribution plots. To solve the above problems, we propose semi-supervised KPIs generation and faulty variable localization methods. Firstly, a physical prior-guided KPIs generation method is proposed for combining unsupervised learning and supervised learning for KPIs under the guidance of physical knowledge. Secondly, a relative contribution plot method for relevant variables is proposed for fault variable localization through the hierarchical correlation calculation of process variables and KPIs. Finally, the experiment validates the monitoring and localization performance of the proposed method in the industrial case.

关键词： Semi-supervised Learning process monitoring Physical Prior Variable Selection

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

Cross-Domain Industrial Process Monitoring Based on Multiple...

process monitoring of the Au-S bond conversion in acetylene hydrochlorination

第35届中国过程控制会议

作者： Ziqing Deng Xiaofang Chen Yongfang Xie Zhaohui Zeng School of Computer Science & School of Cyberspace Science Xiangtan University

Real-word industrial process data typically exhibit varying distributions and suffer from a scarcity of real-time labeled samples, noticeably compromising the performance of process monitoring. To tackle the challenge of inconsistent data distribution between training and testing phases, this paper proposes a multiple structured latent transfer dictionary pair learning(MSLT-DPL) method, applied to cross-domain industrial process monitoring. Specifically, MSLT-DPL leverages knowledge from a source domain with labeled data to monitor a target domain lacking sufficient labeled samples. Through a latent projective double reconstruction strategy, the method integrates transfer learning and latent space mining, effectively bridging distribution disparities between domains. Simultaneously, to boost the discriminative ability, a multiple structured regularization is devised for coding coefficients and significant features. And to enhance the locality in latent space, MSLT-DPL acquires the blockdiagonal locality weights with the designed adaptive locality-preserved function. Subsequently, thorough contrastive experiments confirm the superior performance of the proposed method in cross-domain process monitoring.

关键词： Dictionary learning cross-domain process monitoring projective double reconstruction latent space

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

Chinese Journal of Chemical Engineering 2022年第5期35卷 32-40页

作者： Lutai Song Li Liu Mingyuan Zhu Bin Dai School of Chemistry and Chemical Engineering Shihezi UniversityShihezi 832000China Key Laboratory for Green processing of Chemical Engineering of Xinjiang Bingtuan Shihezi 832000China

The continuous expansion of vinyl chloride production increases environmental pollution caused by mercury catalysts,which is an issue that urgently needs to be *** and stable catalysts should be researched to alleviate this *** this research,Thiolactic acid acts as a ligand where sulfhydryl groups form a stable complex with Au on the surface of a spherical activated carbon(SAC).An Au-thiolactic acid/SAC catalyst was designed with a Au theoretical loading of 0.5%(mass)to overcome the disadvantages of traditional Au-based catalysts,such as a low conversion rate and poor life *** ratio of Au to ligand was screened,and the activity was best when Au/S=1:*** formation of the Au-S bond was proven by FT-IR and *** longevity test of the Au1 S8/SAC catalyst was carried out at 1200 h^(-1) for 50 *** with reaction times of 0 h,5 h,10 h,20 h,and 50 h were taken to monitor the catalyst *** XPS and TPR tests proved that the Au-S bond broke as the acetylene hydrochlorination reaction proceeded,The DFT calculation proved that the Au-S bond is the active site,and the sulfur atom promotes the adsorption of C_(2)H_(2) by the catalyst.

关键词： Acetylene hydrochlorination Au-S bond process monitoring Valence shift

A Novel MDFA-MKECA Method With Application to Industrial Batch process monitoring

在线全文

学校读者我要写书评

暂无评论

IEEE/CAA Journal of Automatica Sinica 2020年第5期7卷 1446-1454页

作者： Yinghua Yang Xiang Shi Xiaozhi Liu Hongru Li College of Information Science and Engineering Northeastern UniversityShenyang 110819China

For the complex batch process with characteristics of unequal batch data length,a novel data-driven batch process monitoring method is proposed based on mixed data features analysis and multi-way kernel entropy component analysis(MDFA-MKECA)in this *** the mechanistic knowledge,different mixed data features of each batch including statistical and thermodynamics entropy features,are extracted to finish data *** that,MKECA is applied to reduce data dimensionality and finally establish a monitoring *** proposed method is applied to a reheating furnace industry process,and the experimental results demonstrate that the MDFA-MKECA method can reduce the calculated amount and effectively provide on-line monitoring of the batch process.

关键词： MDFA MKECA process monitoring reheating furnace statistical features thermodynamics entropy feature

Balanced multiple weighted linear discriminant analysis and its application to visual process monitoring

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Chemical Engineering 2021年第8期34卷 128-137页

作者： Weipeng Lu Xuefeng Yan Key Laboratory of Advanced Control and Optimization for Chemical processes of Ministry of Education East China University of Science and TechnologyShanghai 200237China

Visual process monitoring is important in complex chemical *** address the high state separation of industrial data,we propose a new criterion for feature extraction called balanced multiple weighted linear discriminant analysis(BMWLDA).Then,we combine BMWLDA with self-organizing map(SOM)for visual monitoring of industrial operation *** can extract the discriminative feature vectors from the original industrial data and maximally separate industrial operation states in the space spanned by these discriminative feature *** the discriminative feature vectors are used as the input to SOM,the training result of SOM can differentiate industrial operation states *** function improves the performance of visual *** stirred tank reactor is used to verify that the class separation performance of BMWLDA is more effective than that of traditional linear discriminant analysis,approximate pairwise accuracy criterion,max–min distance analysis,maximum margin criterion,and local Fisher discriminant *** addition,the method that combines BMWLDA with SOM can effectively perform visual process monitoring in real time.

关键词： Linear discriminant analysis process monitoring Self-organizing map Feature extraction Continuous stirred tank reactor process

Fault diagnosis and process monitoring using a statistical pattern framework based on a self-organizing map

在线全文

学校读者我要写书评

暂无评论

Journal of Central South University 2015年第2期22卷 601-609页

作者：宋羽姜庆超颜学峰 Key Laboratory of Advanced Control and Optimization for Chemical processes of Ministry of Education(East China University of Science and Technology)

A multivariate method for fault diagnosis and process monitoring is proposed. This technique is based on a statistical pattern(SP) framework integrated with a self-organizing map(SOM). An SP-based SOM is used as a classifier to distinguish various states on the output map, which can visually monitor abnormal states. A case study of the Tennessee Eastman(TE) process is presented to demonstrate the fault diagnosis and process monitoring performance of the proposed method. Results show that the SP-based SOM method is a visual tool for real-time monitoring and fault diagnosis that can be used in complex chemical *** with other SOM-based methods, the proposed method can more efficiently monitor and diagnose faults.

关键词： statistic pattern framework self-organizing map fault diagnosis process monitoring

Local component based principal component analysis model for multimode process monitoring

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Chemical Engineering 2021年第6期34卷 116-124页

作者： Yuan Li Dongsheng Yang Department of Information Engineering Shenyang University of Chemical TechnologyShenyang 110142China

For plant-wide processes with multiple operating conditions,the multimode feature imposes some challenges to conventional monitoring ***,to solve this problem,this paper provides a novel local component based principal component analysis(LCPCA)approach for monitoring the status of a multimode *** LCPCA,the process prior knowledge of mode division is not required and it purely based on the process ***,LCPCA divides the processes data into multiple local components using finite Gaussian mixture model mixture(FGMM).Then,calculating the posterior probability is applied to determine each sample belonging to which local *** that,the local component information(such as mean and standard deviation)is used to standardize each sample of local ***,the standardized samples of each local component are combined to train PCA monitoring *** on the PCA monitoring model,two monitoring statistics T^(2) and SPE are used for monitoring multimode *** a numerical example and the Tennessee Eastman(TE)process,the monitoring result demonstrates that LCPCA outperformed conventional PCA and LNS-PCA in the fault detection rate.

关键词： Principal component analysis Finite Gaussian mixture model process monitoring Tennessee Eastman(TE)process

Improved process monitoring using the CUSUM and EWMA-based multiscale PCA fault detection framework

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Chemical Engineering 2021年第1期34卷 253-265页

作者： Muhammad Nawaz Abdulhalim Shah Maulud Haslinda Zabiri Syed Ali Ammar Taqvi Alamin Idris Department of Chemical Engineering Universiti Teknologi PETRONAS32610 Bandar Seri IskandarPerakMalaysia Centre of Contaminant Control&Utilization(CenCoU) Universiti Teknologi PETRONAS32610 Bandar Seri IskandarPerakMalaysia Department of Chemical Engineering NED University of Engineering&TechnologyKarachi 75270Pakistan Department of Chemical and Engineering Sciences Karlstad UniversityKarlstadSweden

process monitoring techniques are of paramount importance in the chemical industry to improve both the product quality and plant *** or incipient irregularities may lead to severe degradation in complex chemical processes,and the conventional process monitoring techniques cannot detect these *** this study to improve the performance of monitoring,an online multiscale fault detection approach is proposed by integrating multiscale principal component analysis(MSPCA) with cumulative sum(CUSUM) and exponentially weighted moving average(EWMA) control *** new Hotelling's T~2 and square prediction error(SPE) based fault detection indices are proposed to detect the incipient irregularities in the process *** performance of the proposed fault detection methods was tested for simulated data obtained from the CSTR system and compared to that of conventional PCA and MSPCA based *** results demonstrate that the proposed EWMA based MSPCA fault detection method was successful in detecting the ***,a comparative study shows that the SPEEWMA monitoring index exhibits a better performance with lower values of missed detections ranging from 0% to 0.80% and false alarms ranging from 0% to 21.20%.

关键词： Chemical process system CSTR Fault detection Multiscale Principal component analysis process monitoring

Label propagation dictionary learning based process monitoring method for industrial process with between-mode similarity

在线全文

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2022年第1期65卷 39-55页

作者： Keke HUANG Shijun TAO Yishun LIU Chunhua YANG Weihua GUI School of Automation Central South University

With the development of the industrial cyber-physical systems, a small amount of labeled data and a large amount of unlabeled data are collected from the industrial process. Due to the variation of internal operation conditions and external environment, there is a between-mode similarity between data samples. The scarcity of labeled data and the existence of similarity make it challenging to extract data characteristics. In addition, it creates new challenges to process monitoring. To solve these problems, this study proposes a label propagation dictionary learning method. We first establish the connection between atoms and corresponding profiles and realize the propagation of their labels through graph Laplacian regularization. Then, considering the similarity of samples in the same class, the low-rank constraint is added to sparse coding to strengthen the mutual propagation of labels. Finally, an optimization method is designed to obtain the dictionary and classifier simultaneously. When new data samples arrive, we conduct process monitoring and condition prediction based on the learned dictionary and classifier. Experiments show that the proposed method can achieve satisfactory monitoring performance when compared to several state-of-the-art methods, indicating the superiority of the proposed method.

关键词： process monitoring label propagation dictionary learning between-mode similarity graph Laplacian regularization