The article proposes a systematic method for fault diagnosis and protection from cyberattacks in thermal power generation units, which makes use of the H-infinity Kalman Filter and of the statistical properties of the χ 2 distribution. An approximate linearization procedure is applied first to the the nonlinear state-space description of a power unit that comprises a steam turbine and a synchronous power generator. Using this modelling approach the power unit is synchronized with the grid's frequency. Moreover, an H-infinity Kalman Filter that relies on the approximately linearized model is used for representing the fault-free functioning of the power unit. The measured outputs of the power generation unit (that is the generator's turn angle and speed and the turbine's power) are compared against the estimated outputs which are provided by the H-infinity Kalman Filter. The associated differences form the residuals' sequence which in turn undergoes statistical processing in an aim to solve the power unit's fault diagnosis problem. It is shown that the square of the residuals' vector, suitably weighted by the inverse of the covariance matrix of the measured outputs vector of the power unit follows the χ 2 distribution and stands for a statistical fault detection test. Actually, by exploiting the statistical properties of the χ 2 distribution and its confidence intervals one can define reliable fault thresholds for the power unit's functioning. Whenever the aforementioned fault thresholds are exceeded the existence of a malfunctioning or cyberattack in the power unit can be concluded. Moreover, by applying the previous statistical test to the individual components of the power unit one can distinguish if the fault or cyberattack has taken place in the synchronous generator or in the steam turbine.

H-Infinity kalman filter for condition monitoring of steam-turbine power generation units

Rigatos G.;Siano P.;
2019-01-01

Abstract

The article proposes a systematic method for fault diagnosis and protection from cyberattacks in thermal power generation units, which makes use of the H-infinity Kalman Filter and of the statistical properties of the χ 2 distribution. An approximate linearization procedure is applied first to the the nonlinear state-space description of a power unit that comprises a steam turbine and a synchronous power generator. Using this modelling approach the power unit is synchronized with the grid's frequency. Moreover, an H-infinity Kalman Filter that relies on the approximately linearized model is used for representing the fault-free functioning of the power unit. The measured outputs of the power generation unit (that is the generator's turn angle and speed and the turbine's power) are compared against the estimated outputs which are provided by the H-infinity Kalman Filter. The associated differences form the residuals' sequence which in turn undergoes statistical processing in an aim to solve the power unit's fault diagnosis problem. It is shown that the square of the residuals' vector, suitably weighted by the inverse of the covariance matrix of the measured outputs vector of the power unit follows the χ 2 distribution and stands for a statistical fault detection test. Actually, by exploiting the statistical properties of the χ 2 distribution and its confidence intervals one can define reliable fault thresholds for the power unit's functioning. Whenever the aforementioned fault thresholds are exceeded the existence of a malfunctioning or cyberattack in the power unit can be concluded. Moreover, by applying the previous statistical test to the individual components of the power unit one can distinguish if the fault or cyberattack has taken place in the synchronous generator or in the steam turbine.
2019
978-1-5386-5517-7
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/11386/4726587
 Attenzione

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

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? 0
social impact