Equipment failure prognostics aims at the prediction of the equipment future health condition. For this, a model is trained to fit the monitored data of the condition of the equipment. This can be quite difficult and complicated with large and nonmonotonic data. Peaks in nonmonotonic data are data values which are much larger than the others. For nonmonotonic degradation processes, it can be more efficient to predict the time and value of the peaks during the evolution of the condition of the equipment. In this paper, we propose a framework to predict the values of the future peaks. More specifically, data-driven models are trained to predict the next peak values, based on the information of the peaks in the historical data. The trained models are, then, used to predict the next peak values. A real case study regarding the leakage of the first seal of a reactor coolant pump in a nuclear power plant is considered to verify the effectiveness of the proposed prediction framework.

Prediction of peak values in time series data for prognostics of critical components in nuclear power plants

ZIO, ENRICO
2016-01-01

Abstract

Equipment failure prognostics aims at the prediction of the equipment future health condition. For this, a model is trained to fit the monitored data of the condition of the equipment. This can be quite difficult and complicated with large and nonmonotonic data. Peaks in nonmonotonic data are data values which are much larger than the others. For nonmonotonic degradation processes, it can be more efficient to predict the time and value of the peaks during the evolution of the condition of the equipment. In this paper, we propose a framework to predict the values of the future peaks. More specifically, data-driven models are trained to predict the next peak values, based on the information of the peaks in the historical data. The trained models are, then, used to predict the next peak values. A real case study regarding the leakage of the first seal of a reactor coolant pump in a nuclear power plant is considered to verify the effectiveness of the proposed prediction framework.
2016
IFAC
critical component; peak value; prediction; support vector machine; time series data; Control and Systems Engineering
File in questo prodotto:
Non ci sono file associati a questo prodotto.

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/11311/1020932
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 7
  • ???jsp.display-item.citation.isi??? 7
social impact