Alarm floods represent a widespread issue for modern chemical plants. During these conditions, the number of alarms may be unmanageable, and the operator may miss safety-critical alarms. Chattering alarms, which repeatedly change between the active and non-active states, are responsible for most of the alarm records within a flood episode. Typically, chattering alarms are only addressed and removed retrospectively (e.g. during periodic performance assessments). This study proposes a Machine-Learning based approach for alarm chattering prediction. Specifically, a method for dynamic chattering quantification has been developed, whose results have been used to train three different Machine Learning models – Linear, Deep, and Wide&Deep models. The algorithms have been employed to predict future chattering behavior based on actual plant conditions. Performance metrics have been calculated to assess the correctness of predictions and to compare the performance of the three models.

Predicting chattering alarms: A machine Learning approach

Tamascelli N.;Cozzani V.
2020

Abstract

Alarm floods represent a widespread issue for modern chemical plants. During these conditions, the number of alarms may be unmanageable, and the operator may miss safety-critical alarms. Chattering alarms, which repeatedly change between the active and non-active states, are responsible for most of the alarm records within a flood episode. Typically, chattering alarms are only addressed and removed retrospectively (e.g. during periodic performance assessments). This study proposes a Machine-Learning based approach for alarm chattering prediction. Specifically, a method for dynamic chattering quantification has been developed, whose results have been used to train three different Machine Learning models – Linear, Deep, and Wide&Deep models. The algorithms have been employed to predict future chattering behavior based on actual plant conditions. Performance metrics have been calculated to assess the correctness of predictions and to compare the performance of the three models.
2020
Tamascelli N.; Paltrinieri N.; Cozzani V.
File in questo prodotto:
Eventuali allegati, non sono esposti

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11585/795039
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 16
  • ???jsp.display-item.citation.isi??? 11
social impact