Method of estimation on a curve of a relevant point for the detection of an anomaly of a motor and data processing system for the implementation thereof

• US 9,792,741 B2
• Filed: 07/21/2014
• Issued: 10/17/2017
• Est. Priority Date: 07/23/2013
• Status: Active Grant

• Current Assignee: Safran Aircraft Engine Company
• Original Assignee: Snecma (Safran SA)
• Inventors: Rabenoro, Tsirizo, Lacaille, Jerome Henri Noel
• Primary Examiner(s): Trivedi, Atul
• Application Number:
  US14/906,470
• Publication Number:
  US 20160163132A1
• Time in Patent Office: 1,184 Days
• Field of Search: 701 99
• CPC Class Codes: F01D 21/003 Arrangements for testing or measuring for measuring vibrations G01H
• CPC Class Codes: F02C 7/00 Features, components parts, details or accessories, not provided for in, or of interest apart form groups F02C1/00 - F02C6/00; Air intakes for jet-propulsion plants controlling F02C9/00
• CPC Class Codes: F05D 2260/80 Diagnostics
• CPC Class Codes: F05D 2260/821 Parameter estimation or prediction
• CPC Class Codes: G05B 23/0243 model based detection method, e.g. first-principles knowledge model
• CPC Class Codes: G05B 23/0254 based on a quantitative model, e.g. mathematical relationships between inputs and outputs; functions: observer, Kalman filter, residual calculation, Neural Networks
• CPC Class Codes: G07C 5/0808 Diagnosing performance data testing of vehicles G01M17/00; testing of electrical installation on vehicles G01R31/005
• CPC Class Codes: G07C 5/0825 using optical means
• CPC Class Codes: G07C 5/12 in graphical form

Abstract

A method of estimation on a curve of a relevant point for detecting an anomaly of a motor. The method includes selecting a profile including a binary code, each component of which codes a direction of variation between two consecutive characteristic points of at least one learning curve, a model making it possible to estimate a relevant point based on a set of characteristic points of a curve and a filter. The method also includes applying the filter of the profile selected to the curve, determining a set of characteristic points of the filtered curve and of a binary code, comparing the determined code and the code of the profile selected, and estimating, as a function of the comparison, the relevant point on the curve based on the characteristic points of the filtered curve and the model of the profile selected.

Claims

1. A method for estimating on a curve a relevant point for detecting an anomaly of an engine, said curve representing a time-dependent change in physical operation parameters of the engine measured by at least one sensor on said engine,

applied by a computer connected to first storage means,

said first storage means storing in memory at least one profile comprising a directional binary code, each component of which codes a direction of variation between two consecutive characteristic points of at least one learning curve, a model allowing estimation of a relevant point from a set of characteristic points of a curve, and a filter,

said method comprising:

a/ selecting a profile stored in memory in the first storage means;

b/ applying the filter of the selected profile to said curve;

c/ determining a set of characteristic points of said filtered curve and of the directional binary code, each component of which codes the direction of variation of two consecutive characteristic points belonging to said set of characteristic points;

d/ comparing the determined directional binary code and the code of the selected profile;

e/ depending on said comparison, estimating the relevant point on said curve from the characteristic points of said filtered curve and from the model of the selected profile.

2. The method according to claim 1, wherein if the determined directional binary code is different from the code of the selected profile, a new profile stored in memory in the first storage means is selected and the computer again executes steps b/ to e/.

3. The method according to claim 1, wherein the relevant point is selected from among an instant for opening a valve, an instant for sharp variation of a temperature or pressure, an instant for attaining certain conditions by a high pressure compressor or a low pressure compressor, a disengagement instant of a starter.

4. The method according to claim 1, wherein the characteristic points of curves are selected from among inflection points, local extrema, sudden changes in slopes.

5. The method according to claim 1, wherein a profile further comprises a threshold and wherein the characteristic points are consecutive local extrema, for which the difference in ordinates is greater than said threshold.

6. The method according to claim 1, wherein said models are generalized linear models with selection of variables.

7. The method according to claim 1, wherein said models verify the equation:

t=AX

Wherein

t is the abscissa of the relevant point to be estimated,

A is a line vector containing regression coefficients,

X is a column vector, whose the components are abscissas of the characteristic points and of their transforms.

8. The method according to claim 1, further comprising a step for estimating from estimated relevant points of specific indicators selected for their representativity of the operating condition of the engine.

9. The method according to claim 8, further comprising a step for diagnosing the engine from estimated specific indicators.

10. The method according to claim 8, comprising a step for storing in memory in storage means, flight after flight, estimated specific indicators and a step for predicting a degradation of the operation state of the engine from the time-dependent change of the specific indicators stored in memory.

11. The method according to claim 1, wherein each profile stored in memory in the first storage means is determined by a learning process.

12. The method according to claim 11, wherein the learning process for a profile comprises:

a/ displaying by a display device several learning curves;

b/ determining by said computer a relevant point for detecting an anomaly on each of the learning curves, said relevant point of each of the learning curves being selected by an expert by means of input means;

c/ storing in second storage means each of said learning curves associated with said determined relevant point;

d/ selecting with said computer a filter and a model;

e/ applying with said computer the selected filter to each of the learning curves;

f/ determining with said computer characteristic points of each of the filtered learning curves;

g/ determining with said computer from the determined characteristic points, recurrent characteristic points appearing on each filtered learning curve and a directional binary code, for which each component codes the direction of variation between two consecutive recurrent characteristic points;

h/ from determined recurrent characteristic points and from the selected model, estimating with said computer the relevant point;

i/ estimating with said computer the error associated with the selected filter and model in step d/ by comparing the relevant point estimated with the relevant point selected by the expert in step b/;

j/ storing in memory, in said first storage means, a profile comprising said determined directional binary code, said filter and said model allowing minimization of the estimated error or obtaining an estimated error below a determined threshold.

13. A computer program product comprising program code instructions for executing steps of the method according to claim 1 when said program is executed on a computer.

14. A data processing system comprising a computer, input means, at least one display device configured for applying the steps of the method according to claim 1.