Monitoring Nonlinear and Non-Gaussian Processes Using Gaussian Mixture Model-Based Weighted Kernel Independent Component Analysis

Cai, Ling; Tian, Xuemin; Chen, Sheng

doi:10.1109/tnnls.2015.2505086

Cited by 100 publications

(45 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…And the detailed Matlab Simulink modules are available in http://depts.washington.edu/control/LARRY/TE/download.html. In the present studies, this process has been a benchmark case widely used to evaluate different monitoring strategies . The TE process is composed of 5 major operation units, including the product condenser, the reactor, the compressor, the separator, and the stripper, as shown in Figure .…”

Section: Case Studymentioning

confidence: 90%

“…In the present studies, this process has been a benchmark case widely used to evaluate different monitoring strategies. [38][39][40][41][42] The TE process is composed of 5 major operation units, including the product condenser, the reactor, the compressor, the separator, and the stripper, as shown in Figure 9. According to the G/H mass ratios, 6 operating modes can be generated.…”

Section: Te Processmentioning

confidence: 99%

See 1 more Smart Citation

Multimode complex process monitoring using double‐level local information based local outlier factor method

Wang

Deng

Cao

2018

Journal of Chemometrics

View full text Add to dashboard Cite

Industrial processes typically have multiple operating modes with complex data distribution and locality faults, which challenges the traditional multivariate statistical process monitoring methods. To address this problem, a double‐level local information‐based local outlier factor (LOF) method is proposed in this work for multimode complex process monitoring. First, to handle the multimodality, the local neighborhood standardization strategy is adopted to utilize the statistical information of local data structure. Second, the variable LOF method is proposed to determine reasonable boundary for complex data distribution and simultaneously reflect local variable behaviors. For better online monitoring, a weighting strategy is applied to emphasize the local variable information, and the Bayesian inference is employed to integrate the LOF value of each variable. To isolate the fault variables, a contribution plot is designed. Finally, a numerical example and the benchmark Tennessee Eastman process are used to demonstrate the effectiveness and superiority of the proposed method.

show abstract

Section: Case Studymentioning

confidence: 90%

Section: Te Processmentioning

confidence: 99%

Multimode complex process monitoring using double‐level local information based local outlier factor method

Wang

Deng

Cao

2018

Journal of Chemometrics

View full text Add to dashboard Cite

show abstract

“…3) Apply the Bayesian inference to calculate the posterior fault probabilities using (29) and (30). 4) Calculate the probability-based monitoring statistics P T 2 and P Q using (37) and (38).…”

Section: Offline Model Training Stagementioning

confidence: 99%

Fault discriminant enhanced kernel principal component analysis incorporating prior fault information for monitoring nonlinear processes

Deng

Tian

Chen

et al. 2017

Chemometrics and Intelligent Laboratory Systems

Self Cite

View full text Add to dashboard Cite

Kernel principal component analysis (KPCA) based fault detection method, whose statistical model only utilizes normal operating data and ignores available prior fault information, may not provide the best fault detection performance for nonlinear process monitoring. In order to exploit available prior fault data to enhance fault detection performance, a fault discriminant enhanced KPCA (FDKPCA) method is proposed, which simultaneously monitors two types of data features, nonlinear kernel principal components (KPCs) and fault discriminant components (FDCs). More specifically, based on the normal operating data, KPCs are extracted by usual KPCA modeling, while with the normal operating data and prior fault data, FDCs are obtained by the kernel local-nonlocal preserving discriminant analysis (KLNPDA). Monitoring statistics are constructed for both the KPCA and KLNPDA sub-models. Moreover, Bayesian inference is employed to transform the corresponding monitoring statistics into fault probabilities, and the overall probability-based monitoring statistics are constructed by weighting the results of the two sub-models, which provides more effective on-line fault detection capability. Two case studies involving a simulated nonlinear system and a continuous stirred tank reactor demonstrate the superior fault detection performance of the proposed KLNPDA scheme over the traditional KPCA method.

show abstract

“…At present, MSPC-based techniques are firmly based on core component technologies that predominantly include principal component analysis, partial least squares, independent component analysis and their extensions (Rato, Reis, Schmitt, Hubert & De Ketelaere, 2016;Li, Qin & Zhou, 2010;Lee, Qin & Lee, 2006;Yu, Khan & Garaniya, 2016;Zhang, Sun & Fan, 2015;Li & Yang, 2015;Cai, Tian & Chen, 2017). These techniques, however, rely on the fundamental assumptions that (i) the process operates at a predefined stationary operating condition, where the recorded variables set has a joint multivariate distribution with a time-invariant mean vector and covariance matrix, and (ii) the recorded set possess no serial correlation.…”

Section: Introductionmentioning

confidence: 99%

Monitoring nonstationary and dynamic trends for practical process fault diagnosis

Lin

Krüger

et al. 2019

Control Engineering Practice

View full text Add to dashboard Cite

This article introduces a revised common trend framework to monitor nonstationary and dynamic trends in industrial processes and shows needs for each improvement on the basis of three application studies. These improvements relate to (i) the extension of the common trend framework to include sets that contain stationary and nonstationary variables, (ii) handling cases where residuals are not drawn from multivariate normal distributions and (iii) the application of the framework to larger variable sets. Existing work does not adequately address these practically important issues. Industrial application studies highlight the needs for (i) the extended framework to model data sets containing stationary and nonstationary variables, (ii) handling statistics that are not based on normally distributed residuals and (iii) the use of Chigira procedure to robustly extract common trends. The extended framework is compared to traditional approaches. Keywords common trends models; Kasa decomposition; cointegration; latent stationary and nonstationary factors; generalized data structure; forecast recovery; multivariate normal distribution processes has been presented through the application to a glass melter process (Lin, Kruger & Chen, 2017), the following practically relevant issues have not been considered: the application studies presented herein contain recorded variable sets that are both, nonstationary and stationary, whereas Lin, Kruger & Chen (2017) addressed the problem of variable sets that are nonstationary only;  the model residuals may not be drawn from multivariate normal distributions, in contrast to the work in Lin, Kruger & Chen (2017) where this assumption is made; and  the conventional cointegration testing procedure by Johansen (1995) may run into difficulties if the number of variables is large, which is typically the case for complex industrial processes.This article is divided into the following sections. Section 2 gives a detailed description of the processes, including the FCCU simulator, the polymerization and the distillation process, used in this article. Section 3 then summarizes the monitoring framework and introduces an improved common trends framework. Sections 4 to 6 summarize the applications to the three processes. A concluding summary is given in Section 7. Process DescriptionsThis section provides a detailed description of the three processes that are used to demonstrate the working of the improved cointegration framework. Subsection 2.1 gives a description of the simulated fluid catalytic cracking unit. Subsections 2.2 and 2.3 then provide details of the polymerization and distillation processes, respectively. Fluid Catalytic Cracking UnitA fluid catalytic cracking unit (FCCU) is an important part of a refinery. An FCCU receives several different heavy feed stocks from other refinery units and cracks these into more valuable components that are usually blended into gasoline and other products. The main feed stream to an FCCU is gas oil, but heavier diesel and wash oil streams also cont...

show abstract

Monitoring Nonlinear and Non-Gaussian Processes Using Gaussian Mixture Model-Based Weighted Kernel Independent Component Analysis

Cited by 100 publications

References 36 publications

Multimode complex process monitoring using double‐level local information based local outlier factor method

Multimode complex process monitoring using double‐level local information based local outlier factor method

Fault discriminant enhanced kernel principal component analysis incorporating prior fault information for monitoring nonlinear processes

Monitoring nonstationary and dynamic trends for practical process fault diagnosis

Contact Info

Product

Resources

About