Standardizing data used for monitoring an aeroengine

1. A method of standardizing data used for monitoring an aeroengine, the method comprising:

collecting time-series measurements over time concerning said aeroengine;

from said time-series measurements, calculating a set of indicators Y=(y 1 , . . . , y j , . . . , y m ) that are specific to elements of said engine;

from said time-series measurements, identifying an exogenous data set X=(x 1 , . . . , x n ), representative of external context acting on said set of indicators Y;

for each indicator y j of said set of indicators Y, constructing a projection space E (j) =σ(Y (j) , X) generated by analytic transformations of a subset of indicators Y (j) =(y 1 , . . . y j−1 , y j+1 , . . . y m ) comprising all of the indicators of said set of indicators Y except each said indicator y j and by said exogenous data set X;

for each indicator y j of said set of indicators Y, calculating a corresponding estimator ŷ j by projecting said indicator y j using a regression technique onto said projection space E (j) =σ(Y (j) , X), thereby forming a set of estimators Ŷ=(ŷ 1 , . . . , ŷ j , . . . , ŷ m ) corresponding to said set of indicators Y=(y 1 , . . . , y j , . . . , y m ); and

normalizing each estimator ŷ j as a function of a reference value for the corresponding indicator y j and of a difference between each said estimator ŷ j and said corresponding indicator y j to form a set of standardized values {tilde over (Y)}=({tilde over (y)} 1 , . . . {tilde over (y)} j , . . . , {tilde over (y)} m ).

2. A method according to claim 1 , wherein said time-series measurements are collected during normal operation of said aeroengine.

3. A method according to claim 1 , wherein each standardized value {tilde over (y)} j is calculated by adding a mean or reference value for the corresponding indicator y j to a difference between the corresponding indicator y j and the corresponding estimator ŷ j , using the following equation: {tilde over (y)} j = y j +(y j −ŷ j ).

4. A method according to claim 1 , further comprising analyzing robustness of each estimator using a cross evaluation technique serving to select an optimum projection space.

5. A method according to claim 1 , wherein said projection space is constructed using expert criteria with help of physical formulations of relationships between the indicators and between the indicators and the exogenous data.

6. A method according to claim 1 , wherein said space is constructed automatically by using a neural network.

7. A method according to claim 6 , wherein said neural network is a model having nodes.

8. A method according to claim 1 , wherein the indicators of said set of indicators Y=(y 1 , . . . , y j , . . . , y m ) are specific to physical and/or logical elements of said engine.

9. A method according to claim 1 , wherein the indicators of said set of indicators Y=(y 1 , . . . , y j , . . . , y m ) are calculated using expert criteria by constructing an FMECA.

10. A method according to claim 1 , wherein the indicators are identified by referencing particular points or particular functions summarizing details or shapes of certain curves representative of said time-series measurements.

11. A method according to claim 1 , wherein the exogenous data X=(x 1 , . . . , x n ) acting on the indicators is identified using expert criteria by dependency analysis enabling context data associated with the indicators to be listed.

12. A method according to claim 1 , wherein said regression is a linear regression.

13. A system for standardizing data used for monitoring an aeroengine, the system comprising:

means for operating over time to collect time-series measurements from said aeroengine;

means for calculating from said time-series measurements a set of indicators Y=(y 1 , . . . y j , . . . , y m ) specific to elements of said engine;

means for identifying from said time-series measurements an exogenous data set X=(x 1 , . . . , x n ) representative of external context acting on said set of indicators Y;

means for constructing for each indicator y j of said set of indicators Y, a projection space E (j) =σ(Y (j) , X) generated by analytic transformations of a subset of indicators Y (j) =(y 1 , . . . , y j−1 , y j+1 , . . . , y m ) comprising all of the indicators of said set of indicators Y except each said indicator y j , and by said exogenous data set X;

means for calculating for each said indicator y j of said set of indicators Y a corresponding estimator ŷ j by using a regression technique to project each said indicator y j onto said projection space E (j) =σ(Y (j) , X), forming a set of estimators Ŷ=(ŷ 1 , . . . , ŷ j , . . . , ŷ m ) corresponding to said set of indicators Y=(y 1 , . . . , y j , . . . , y m ); and

means for normalizing each estimator ŷ j as a function of a reference value for the corresponding indicator y j and of a difference between each said estimator ŷ j and said corresponding indicator y j so as to form a set of standardized values {tilde over (Y)}=({tilde over (y)} 1 , . . . , {tilde over (y)} j , . . . , {tilde over (y)} m ).

14. A non-transitory computer readable medium including computer executable instructions for implementing the standardization method according to claim 1 when executed by a processor.