Publication :
Development of a soft sensor for detecting overpitched anodes : detailed investigation of an anode sticking event

En cours de chargement...
Vignette d'image
Date
2021-10-15
Direction de publication
Direction de recherche
Titre de la revue
ISSN de la revue
Titre du volume
Éditeur
Springer
Projets de recherche
Structures organisationnelles
Numéro de revue
Résumé
Adjusting pitch ratio in green anode formulation is becoming difficult due to the increasing raw material variability. The optimal quantity of pitch yielding the best anode properties for a given aggregate, known as the optimal pitch demand (ODP), changes more frequently and is unknown a priori. Exceeding the OPD increases the risk of generating post-baking anode-sticking events. Previously, the potential of a principal component analysis (PCA)-based monitoring scheme for detecting the onset of these undesirable events was assessed by using a set of five green anode resistivity measurements collected from over 120,000 anodes produced over a two-year period. The squared prediction error (SPE) was shown to be sensitive to abnormal events such as anode sticking. The objective of this paper is to further validate the soft sensor by studying the SPE dynamic behavior during a post-baking sticking event when changes in the anode paste formulation were introduced as part of normal operation. Descriptive and statistical analyses demonstrate that the SPE metric reacts significantly to changes in the recipe. The provided example illustrates how the SPE metric used together with pitch ratio data could help advised the operators of manufacturing conditions posing a higher risk of generating post-baking sticking problems.
Description
Revue
Metallurgical and Materials Transactions B, Vol. 52, 4157-4166 (2021)
DOI
10.1007/s11663-021-02335-y
URL vers la version publiée
Mots-clés
Anode sticking , Principal Component Analysis , Squared Prediction Error , Green anode formulation , Resistivity measurement , Optimum pitch demand
Citation
Type de document
article de recherche