Documents disponibles écrits par cet auteur

Multivariate temporal data analysis using self-organizing maps. 1. Training methodology for effective visualization of multistate operations / Yew Seng Ng in Industrial & engineering chemistry research, Vol. 47 N°20 (Octobre 2008) 

Public ISBD [article]  in Industrial & engineering chemistry research > Vol. 47 N°20 (Octobre 2008) . - p. 7744-7757 Titre : Multivariate temporal data analysis using self-organizing maps. 1. Training methodology for effective visualization of multistate operations Type de document : texte imprimé Auteurs : Yew Seng Ng, Auteur ; Rajagopalan Srinivasan, Auteur Année de publication : 2008 Article en page(s) : p. 7744-7757 Note générale : Chemical engineering Langues : Anglais (eng) Mots-clés : Self-organizing  map  (SOM)  Multivariate  temporal  data  analysis Résumé : Multistate operations are common in chemical plants and result in high-dimensional, multivariate, temporal data. In this two-part paper, we develop self-organizing map (SOM)-based approaches for visualizing and analyzing such data. In Part 1 of this paper, the SOM is used to reduce the dimensionality of the data and effectively visualize multistate operations in a three-dimensional map. A key characteristic of multistate processes is that the plant operates for long durations at steady states and undergoes brief transitions involving large changes in variable values. When classical SOM training algorithms are used on data from multistate processes, large portions of the SOM become dedicated to steady states, which exaggerates even minor noise in the data. Also, transitions are represented as discrete jumps on the SOM space, which makes it an ineffective tool for visualizing multistate operations. In this Part 1, we propose a new training strategy specifically targeted at multistate operations. In the proposed strategy, the training dataset is first resampled to yield equal representation of the different process states. The SOM is trained with this state-sampled dataset. Furthermore, clustering is applied to group neurons of high similarity into compact clusters. Through this strategy, modes and transitions of multistate operations are depicted differently, with process modes visualized as intuitive clusters and transitions as trajectories across the SOM. We illustrate the proposed strategy using two real-case studies, namely, startup of a laboratory-scale distillation unit and operation of a refinery hydrocracker. En ligne : http://pubs.acs.org/doi/abs/10.1021/ie0710216 [article] Multivariate temporal data analysis using self-organizing maps. 1. Training methodology for effective visualization of multistate operations [texte imprimé] / Yew Seng Ng, Auteur ; Rajagopalan Srinivasan, Auteur . - 2008 . - p. 7744-7757. Chemical engineering Langues : Anglais (eng) in Industrial & engineering chemistry research > Vol. 47 N°20 (Octobre 2008) . - p. 7744-7757 Mots-clés : Self-organizing  map  (SOM)  Multivariate  temporal  data  analysis Résumé : Multistate operations are common in chemical plants and result in high-dimensional, multivariate, temporal data. In this two-part paper, we develop self-organizing map (SOM)-based approaches for visualizing and analyzing such data. In Part 1 of this paper, the SOM is used to reduce the dimensionality of the data and effectively visualize multistate operations in a three-dimensional map. A key characteristic of multistate processes is that the plant operates for long durations at steady states and undergoes brief transitions involving large changes in variable values. When classical SOM training algorithms are used on data from multistate processes, large portions of the SOM become dedicated to steady states, which exaggerates even minor noise in the data. Also, transitions are represented as discrete jumps on the SOM space, which makes it an ineffective tool for visualizing multistate operations. In this Part 1, we propose a new training strategy specifically targeted at multistate operations. In the proposed strategy, the training dataset is first resampled to yield equal representation of the different process states. The SOM is trained with this state-sampled dataset. Furthermore, clustering is applied to group neurons of high similarity into compact clusters. Through this strategy, modes and transitions of multistate operations are depicted differently, with process modes visualized as intuitive clusters and transitions as trajectories across the SOM. We illustrate the proposed strategy using two real-case studies, namely, startup of a laboratory-scale distillation unit and operation of a refinery hydrocracker. En ligne : http://pubs.acs.org/doi/abs/10.1021/ie0710216

Multivariate temporal data analysis using self-organizing maps. 2. Monitoring and diagnosis of multistate operations / Yew Seng Ng in Industrial & engineering chemistry research, Vol. 47 N°20 (Octobre 2008) 

Public ISBD [article]  in Industrial & engineering chemistry research > Vol. 47 N°20 (Octobre 2008) . - P. 7758-7771 Titre : Multivariate temporal data analysis using self-organizing maps. 2. Monitoring and diagnosis of multistate operations Type de document : texte imprimé Auteurs : Yew Seng Ng, Auteur ; Rajagopalan Srinivasan, Auteur Année de publication : 2008 Article en page(s) : P. 7758-7771 Note générale : Chemical engineering Langues : Anglais (eng) Mots-clés : Self-organizing  map  (SOM)  Multivariate  temporal  data  analysis Résumé : The operation of transitions in continuous processes is challenging and often results in out-of-spec products, alarm floods, and abnormal situations. Therefore, efficient techniques for automated monitoring and fault diagnosis of such operations are essential. In Part 1 of this paper, we proposed a self-organizing map (SOM) training strategy for effectively visualizing multistate operations. In this part of the series, we use the same methodology as a representation scheme to compare operating trajectories and diagnosing faults during transient operations. In the proposed approach, clusters of SOM neurons, called neuronal clusters, serve as landmarks on the multivariate measurement space. Online data during the transition are reflected as a trajectory on the SOM and are converted to a sequence of neuronal clusters, which are the signature of the operating state. We have adapted the well-known Smith and Waterman discrete sequence comparison algorithm from bioinformatics to compare the state signatures and account for run-to-run variations. The proposed comparison method accounts explicitly for oscillations that are common in chemical processes. Online monitoring and diagnosis is performed by comparing the signature with those of known normal and abnormal transitions. The key advantage of the proposed strategy are its computational speed, inherent multivariate nature, and robustness to run-to-run variations, in addition to intuitiveness and visualization of the results. We illustrate the proposed method through two case studies: the Tennessee Eastman challenge problem and startup of a laboratory-scale distillation unit. En ligne : http://pubs.acs.org/doi/abs/10.1021/ie071022y [article] Multivariate temporal data analysis using self-organizing maps. 2. Monitoring and diagnosis of multistate operations [texte imprimé] / Yew Seng Ng, Auteur ; Rajagopalan Srinivasan, Auteur . - 2008 . - P. 7758-7771. Chemical engineering Langues : Anglais (eng) in Industrial & engineering chemistry research > Vol. 47 N°20 (Octobre 2008) . - P. 7758-7771 Mots-clés : Self-organizing  map  (SOM)  Multivariate  temporal  data  analysis Résumé : The operation of transitions in continuous processes is challenging and often results in out-of-spec products, alarm floods, and abnormal situations. Therefore, efficient techniques for automated monitoring and fault diagnosis of such operations are essential. In Part 1 of this paper, we proposed a self-organizing map (SOM) training strategy for effectively visualizing multistate operations. In this part of the series, we use the same methodology as a representation scheme to compare operating trajectories and diagnosing faults during transient operations. In the proposed approach, clusters of SOM neurons, called neuronal clusters, serve as landmarks on the multivariate measurement space. Online data during the transition are reflected as a trajectory on the SOM and are converted to a sequence of neuronal clusters, which are the signature of the operating state. We have adapted the well-known Smith and Waterman discrete sequence comparison algorithm from bioinformatics to compare the state signatures and account for run-to-run variations. The proposed comparison method accounts explicitly for oscillations that are common in chemical processes. Online monitoring and diagnosis is performed by comparing the signature with those of known normal and abnormal transitions. The key advantage of the proposed strategy are its computational speed, inherent multivariate nature, and robustness to run-to-run variations, in addition to intuitiveness and visualization of the results. We illustrate the proposed method through two case studies: the Tennessee Eastman challenge problem and startup of a laboratory-scale distillation unit. En ligne : http://pubs.acs.org/doi/abs/10.1021/ie071022y