Detection of Plant-Wide Disturbances Using a Spectral Classification Tree

Thornhill, Nina F.; Melbø, Hallgeir

doi:10.3182/20060402-4-br-2902.01145

Cited by 2 publications

(3 citation statements)

References 29 publications

(20 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thornhill, Shah, Huang and Visnubhotla [2002] used spectral PCA to find clusters of measurements having similar spectra, detecting clusters of disturbances both with distinct spectral peaks and with multiple spectral features. Methods for display include a colour map [Tangirala, Shah & Thornhill, 2005] or a hierarchical tree [Thornhill and Melbø, 2006].…”

Section: Detection Of Multiple Oscillations and Nonoscillating Disturmentioning

confidence: 99%

Advances and new directions in plant-wide disturbance detection and diagnosis

Thornhill

Horch²

2007

Control Engineering Practice

Self Cite

169

110

View full text Add to dashboard Cite

Section: Detection Of Multiple Oscillations and Nonoscillating Disturmentioning

confidence: 99%

Advances and new directions in plant-wide disturbance detection and diagnosis

Thornhill

Horch²

2007

Control Engineering Practice

Self Cite

169

110

View full text Add to dashboard Cite

“…Related works on enhancing the visualization of multidimensional principal components has been reported by Wang et al 26 and Thornhill and Melbo. 27 In contrast, the need for multidimensional visualization is eliminated by NMF, because it obtains a set of close-to non-overlapping decomposition. Consequently, it is sufficient to look at each component individually.…”

Section: Isolation Of Loops With Commonmentioning

confidence: 99%

“…In fact, these limitations become even more stringent as the required number of principal components becomes large. Related works on enhancing the visualization of multidimensional principal components has been reported by Wang et al and Thornhill and Melbo . In contrast, the need for multidimensional visualization is eliminated by NMF, because it obtains a set of close-to non-overlapping decomposition.…”

Section: Industrial Case Studiesmentioning

confidence: 99%

Non-Negative Matrix Factorization for Detection and Diagnosis of Plantwide Oscillations

Tangirala

Kanodia

Shah

2007

Ind. Eng. Chem. Res.

View full text Add to dashboard Cite

In this paper, we propose the use of non-negative matrix factorization (NMF) of multivariate spectra for plantwide oscillation detection. One of the key features of NMF is that it provides a parts-based representation that allows us to retain the causal basis spectral shapes or parts that constitute the spectra of measurements, unlike the popular principal component analysis (PCA)-based methods. The contributions of this paper are as follows: (i) a novel measure known as the pseudo-singular value (PSV) to assess the order of the basis space (the PSV is also useful in determining the most dominant features of a data set); (ii) a power decomposition plot that contains the total power (defined in this work) and its decomposition by NMF (the power plot is a useful and compact visual tool that provides overall spectral characteristics of the plant and shows the decomposition of these characteristics into well-localized frequency components); and (iii) a novel measure defined as the strength factor (SF) to assess the strength of the localized features in the variables (it can be also used in isolating the root cause). Finally, it is shown that the proposed implementation of NMF is powerful and sensitive enough to capture small oscillations in the measurements. As a result, it largely eliminates the need to filter the data. Industrial case studies are presented to illustrate the applications of NMF and to demonstrate the utility and practicality of the proposed measures.

show abstract

Detection of Plant-Wide Disturbances Using a Spectral Classification Tree

Cited by 2 publications

References 29 publications

Advances and new directions in plant-wide disturbance detection and diagnosis

Advances and new directions in plant-wide disturbance detection and diagnosis

Non-Negative Matrix Factorization for Detection and Diagnosis of Plantwide Oscillations

Contact Info

Product

Resources

About