Kernel PLS-based GLRT method for fault detection of chemical processes

Botre, Chiranjivi; Mansouri, Majdi; Nounou, Mohamed; Nounou, Hazem; Karim, M. Nazmul

doi:10.1016/j.jlp.2016.05.023

Cited by 102 publications

(37 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The works of Jiang and Yan [143,144] improved the sensitivity of kernel PCA by investigating the rate of change of the statistical index and by giving a weight to each feature. Lastly, a new statistic based on the generalized likelihood ratio test (GLRT) can also improve detection for kernel PCA and kernel PLS, as shown by Mansouri et al [192,193,210,270,271].…”

Section: Improved Sensitivity and Incipient Fault Detectionmentioning

confidence: 99%

A Review of Kernel Methods for Feature Extraction in Nonlinear Process Monitoring

et al. 2019

View full text Add to dashboard Cite

Kernel methods are a class of learning machines for the fast recognition of nonlinear patterns in any data set. In this paper, the applications of kernel methods for feature extraction in industrial process monitoring are systematically reviewed. First, we describe the reasons for using kernel methods and contextualize them among other machine learning tools. Second, by reviewing a total of 230 papers, this work has identified 12 major issues surrounding the use of kernel methods for nonlinear feature extraction. Each issue was discussed as to why they are important and how they were addressed through the years by many researchers. We also present a breakdown of the commonly used kernel functions, parameter selection routes, and case studies. Lastly, this review provides an outlook into the future of kernel-based process monitoring, which can hopefully instigate more advanced yet practical solutions in the process industries.

show abstract

Section: Improved Sensitivity and Incipient Fault Detectionmentioning

confidence: 99%

A Review of Kernel Methods for Feature Extraction in Nonlinear Process Monitoring

et al. 2019

View full text Add to dashboard Cite

show abstract

“…These methods include Shewhart chart, EWMA chart, cumulative sum chart, and GLRT chart . In previous studies, PCA‐based GLRT, PLS‐based GLRT, kernel PCA–based GLRT, and kernel PLS–based GLRT techniques have been investigated for failure detection in different applications. In Hu et al, a failure diagnosis strategy is realized using parameter‐based PV model obtained by combining energy balance equation and electrical model.…”

Section: Related Work On Failure Detection In Pv Systemsmentioning

confidence: 99%

“…The multivariate detection techniques are mainly based on partial least squares (PLS) and principal component analysis (PCA), while the univariate techniques include cumulative sum, Shewhart, and exponentially weighted moving average (EWMA) charts . In previous studies, the authors have proved that the classical GLRT chart–based equal variances present a better detection efficiency against the EWMA and Shewhart statistics, which is due to the ability of the GLRT to minimize the false alarm rate (FAR).…”

Section: Introductionmentioning

confidence: 99%

Enhanced generalized likelihood ratio test for failure detection in photovoltaic systems

Mansouri

Hajji

Trabelsi

et al. 2018

Int Trans Electr Energ Syst

Self Cite

View full text Add to dashboard Cite

Summary In this paper, a new multiscale weighted generalized likelihood ratio test (MS‐WGLRT) chart is proposed for enhanced failure detection in photovoltaic systems. The main weakness of the classical generalized likelihood ratio test chart is in dealing with residual samples while ignoring their natural variances. By taking into consideration the nature variance of the detection residual and applying a multiscale representation, the proposed technique allows the reduction in false alarm and missed detection rates compared with the classical generalized likelihood ratio test chart. The multiscale representation of data is an efficient data analysis and feature extraction tool that has a great impact on the effectiveness of failure detection. The effectiveness of the proposed method is evaluated on a simulated photovoltaic data where the developed chart is used for detecting single and multiple failures (eg, bypass, mix, and shading failures). The simulation results show that the multiscale weighted generalized likelihood ratio test method offers better performance compared with the classical generalized likelihood ratio chart.

show abstract

“…Notably, ICA‐based linear projection cannot effectively represent data with a nonlinear structure . Many methods are available to deal with nonlinear monitoring problems, such as neural network, the kernel learning method, and the probability method . However, as modern industrial processes have multiple operation units and become more complex, fault detection based on a single model remains difficult …”

Section: Introductionmentioning

confidence: 99%

Process monitoring method based on correlation variable classification and vine copula

Cui

2020

Can J Chem Eng

View full text Add to dashboard Cite

Chemical processes are becoming increasingly complicated, leading to an increase in process variables and more complex relationships among them. The vine copula has a significant advantage in portraying the dependence of high-dimensional variables. However, as the dimensions increase, the vine copula model incurs a high computational load; such pressure greatly reduces model efficiency. Relationships among variables in the industrial process are complex. Different variables may be strongly or weakly associated or even independent. This paper proposes a process monitoring method based on correlation variable classification and vine copula. The weighted correlation measure is first used to divide variables into a correlated subspace and weakly correlated subspace. Then, two vine structures, C-vine and D-vine, are applied to the correlated and weakly correlated subspaces, respectively. This method takes advantage of C-vine for correlated variables and the flexibility of D-vine for weakly correlated variables. Finally, comprehensive statistics are established based on different subspaces. Monitoring results of the numerical system and the Tennessee Eastman process demonstrate the effectiveness and validity of the proposed method.

show abstract

Kernel PLS-based GLRT method for fault detection of chemical processes

Cited by 102 publications

References 20 publications

A Review of Kernel Methods for Feature Extraction in Nonlinear Process Monitoring

A Review of Kernel Methods for Feature Extraction in Nonlinear Process Monitoring

Enhanced generalized likelihood ratio test for failure detection in photovoltaic systems

Process monitoring method based on correlation variable classification and vine copula

Contact Info

Product

Resources

About