Adaptive algorithm-based engine health prediction

1. A method of reflecting recent corrected engine performance in a data storage device on an aircraft, the method comprising,

by a computer comprising a processor and memory:

periodically receiving, from an engine sensor coupled to the computer, engine operating information for the engine;

for each engine operating information that is periodically received, updating a stored data set in the data storage device to reflect a recent performance fingerprint of the engine, the updating comprising:

determining, from the engine operating information, a power parameter and a critical parameter;

correcting the power parameter and the critical parameter based on additional information available in an aircraft data bus on the aircraft;

recording as a data point in the stored data set, the corrected power parameter and the corrected critical parameter;

grouping the data point with statistically similar historical data points in the stored data set;

merging the data point and the statistically similar historical data points into a single representative data point in the stored data set, the stored data set including previously recorded power parameter and critical parameter data points; and

purging one or more older data points in the stored data set.

2. The method according to claim 1 , wherein receiving the engine operating information includes receiving the engine operating information from the aircraft data bus.

3. The method according to claim 1 , wherein the critical parameter is an MGT value.

4. The method according to claim 1 , wherein the power parameter is a power margin.

5. The method according to claim 1 , wherein the data point represents the engine's power if it were performing on a test stand under standard conditions.

6. The method according to claim 1 , further comprising:

determining whether the power parameter and the critical parameter represent a stable data point; and

discarding the engine operating information in the event that the power parameter and the critical parameter do not represent a stable data point.

7. The method according to claim 1 , further comprising:

determining whether the power parameter and the critical parameter represent a valid data point; and

discarding the engine operating information in the event that the power parameter and the critical parameter do not represent a valid data point.

8. The method according to claim 7 , wherein the determining whether the power parameter and the critical parameter represent a valid data point comprises determining whether there is weight on landing gear.

9. The method according to claim 1 , wherein the correcting comprises applying a correction factor based on an outside air temperature.

10. A method of continuously indicating engine health on an aircraft, the method comprising,

by a computer comprising a processor and memory:

periodically receiving, from an engine sensor coupled to the computer, engine operating information for an engine on the aircraft;

for each engine operating information that is periodically received, updating a stored data set in a data storage device on the aircraft to reflect a recent performance fingerprint of the engine, the updating comprising:

recording a data point in the stored data set;

grouping the data point with statistically similar historical data points in the stored data set;

merging the data point and the statistically similar historical data points into a single representative data point in the stored data set; and

purging one or more older data points in the stored data set;

responsive to the updating, determining a health of the engine, the determining comprising:

selecting an engine performance curve to use in creating an engine health model;

executing a curve-fitting process to obtain a fitted curve, the fitted curve being based on the selected engine performance curve and the stored data set;

evaluating engine health by comparing the fitted curve to a 100% specification-level performance curve for the engine, the evaluating yielding engine health information; and

outputting the engine health information to an on-board display.

11. The method according to claim 10 , wherein the selected engine performance curve includes the 100% specification-level performance curve for the engine.

12. The method according to claim 10 , wherein executing a curve-fitting process to obtain a fitted curve includes translating the selected engine performance curve to achieve a low error with respect to the stored data set.

13. The method according to claim 10 , wherein executing a curve-fitting process to obtain a fitted curve includes rotating the selected engine performance curve to achieve a low error with respect to the stored data set.

14. The method according to claim 10 , wherein executing a curve-fitting process to obtain a fitted curve comprises:

translating the selected engine performance curve to achieve a low error with respect to the stored data set; and

rotating the selected engine performance curve to achieve a low error with respect to the stored data set,

wherein the translating and rotating steps are achieved based on an optical fit.

15. The method according to claim 10 , wherein evaluating engine health by comparing the fitted curve to a 100% specification-level performance curve for the engine includes evaluating margins between the fitted curve and the 100% specification-level performance curve at various points to determine an overall estimate of engine health.

16. The method according to claim 10 , wherein evaluating engine health by comparing the fitted curve to a 100% specification-level performance curve for the engine includes evaluating a margin between the fitted curve and the 100% specification-level performance curve at a chosen point to determine an expected power margin at the chosen point.

17. The method according to claim 10 , wherein the determining the health of the engine is performed periodically based on flight hours.

18. A computer-program product comprising a non-transitory computer-usable medium having computer-readable program code embodied therein, the computer-readable program code adapted to be executed to implement a method comprising:

periodically receiving, from an engine sensor coupled to a computer, engine operating information for an engine on an aircraft;

for each engine operating information that is periodically received, updating a stored data set in a data storage device on the aircraft to reflect a recent performance fingerprint of the engine, the updating comprising:

determining, from the engine operating information, a power parameter and a critical parameter;

correcting the power parameter and the critical parameter based on additional information available in an aircraft data bus on the aircraft;

recording, as a data point in the stored data set, the corrected power parameter and the corrected critical parameter;

grouping the data point with statistically similar historical data points in the stored data set;

merging the data point and the statistically similar historical data points into a single representative data point in the stored data set, the stored data set including previously recorded power parameter and critical parameter data points; and

purging one or more older data points in the stored data set; responsive to the updating, determining a health of the engine, the determining comprising:

selecting an engine performance curve to use in creating an engine health model;

obtaining a fitted curve by translating and rotating the selected engine performance curve to achieve a low error with respect to the stored data set;

evaluating engine health by comparing the fitted curve to a 100% specification-level performance curve for the engine, the evaluating yielding engine health information; and

outputting engine health information to an on-board display.