Method for detecting a malfunction suffered by an integrated drive generator, in an aircraft

Patent metadata

Field	Value
Publication number	US-12429512-B2
Application number	US-202318469268-A
Country	US
Kind code	B2
Filing date	Sep 18, 2023
Priority date	Sep 19, 2022
Publication date	Sep 30, 2025
Grant date	Sep 30, 2025

Abstract

A data analysis system comprises an artificial intelligence of the machine learning type. The data analysis system collects reference data relating to measurements made and recorded during previous flights of aircraft. A training of the artificial intelligence of the machine learning type is implemented by virtue of reference data, with a first classification run for training to detect potential malfunctions suffered by integrated drive generators and a second classification run for training to determine causes of the malfunctions, where relevant. After using in production, the data analysis system collects data to be analyzed relating to measurements made and recorded during flights of the aircraft comprising an integrated drive generator to be monitored, and uses the artificial intelligence of the machine learning type to predict a potential malfunction and, where relevant, predict its cause using data to be analyzed.

First claim

The invention claimed is:

1. A method for detecting a malfunction suffered by an integrated drive generator (IDG) in an aircraft, the method being implemented by a data analysis system comprising a machine learning artificial intelligence, wherein the method comprises the following steps:
collecting reference data relating to measurements made and recorded during previous flights of aircraft equipped with IDGs, called reference IDGs, the reference data comprising at least measurements corresponding to the following parameters, for each reference IDG:
a flow of fuel supplying a jet engine with which the respective reference IDG is associated;
a position of a fuel return valve associated with the respective reference IDG;
a temperature measured by a sensor on an inlet port of the respective reference IDG;
a temperature measured by a sensor on an outlet port of the respective reference IDG;
a frequency of electrical signal produced by the respective reference IDG;
a temperature of oil supplying the jet engine with which the respective reference IDG is associated;

carrying out a training, in supervisor mode, of a machine learning artificial intelligence with the reference data, with a first classification run for training to detect potential malfunctions suffered by IDGs and a second classification run for training to determine causes of the malfunctions where relevant;
using in production the machine learning artificial intelligence after validation of the training;
collecting data to be analyzed, the data the same type as the reference data, relating to measurements made and recorded during flights of the aircraft comprising the IDG; and,
using the machine learning artificial intelligence that has been trained to predict a potential malfunction suffered by the IDG and predict, where relevant, a cause of the malfunctions by virtue, respectively, of the said first and second classification runs, using the data to be analyzed.

2. The method according to claim 1 , wherein the reference data further comprise measurements corresponding to at least one of the following parameters, for each reference IDG:
a time during which the fuel return valve is open;
a time during which the fuel return valve is closed;
a ratio between the time during which the fuel return valve is closed and the time during which the fuel return valve is open;
a time taken by the fuel return valve to go from a closed position to an open position; and
a difference in temperature between the inlet port of the respective reference IDG and the outlet port of the respective reference IDG.

3. The method according to claim 1 , wherein the data analysis system processes the reference data and to the data to be analyzed by, at least, calculating the following information, for one or more of the said parameters:
a minimum value;
a maximum value;
a standard deviation;
an average;
a moving average;
a gradient coefficient; or
any combination thereof.

4. The method according to claim 3 , wherein the calculating of at least a part of the said information is carried out by grouping reference data and, respectively, data to be analyzed, over several consecutive aircraft flights.

5. The method according to claim 1 , wherein the machine learning artificial intelligence is trained to detect an IDG malfunction linked to:
an anomaly in the opening/closing controls of a fuel return valve;
a clogging of IDG oil;
a failure of a temperature sensor on the outlet port;
a failure of a temperature sensor on the inlet port;
an oil leak in an IDG;
an anomaly of IDG maintenance;
or any combination thereof.

6. The method according to claim 1 , wherein the first classification run comprises applying a first algorithm and the second classification run comprises applying a second algorithm distinct from the first algorithm.

7. The method according to claim 6 , wherein the first algorithm comprises an LightGBM algorithm and the second algorithm comprises an XGBooost algorithm.

8. The method according to claim 1 , wherein the machine learning artificial intelligence is used in production in an avionics system of the aircraft equipped with the IDG.

9. The method according to claim 1 , wherein the machine learning artificial intelligence is used in production in a computer system on the ground in a control center, and wherein the data to be analyzed are transmitted by an avionics system of the aircraft, equipped with the IDG, to the computer system in the control center.

10. A non-transitory computer readable storage medium comprising:
a computer program comprising instructions which, when executed by a processor, performs the method according to claim 1 .

11. A system comprising:
an electronic circuitry configured to carry out a detection of a malfunction suffered by an integrated drive generatot (IDG) to be monitored, in an aircraft, wherein the electronic circuitry is configured to:
collect reference data relating to measurements made and recorded during previous flights of aircraft equipped with reference IDGs, the reference data comprising at least measurements corresponding to the following parameters, for each reference IDG:
a flow of fuel supplying a jet engine with which the respective reference IDG is associated;
a position of a fuel return valve associated with the respective reference IDG;
a temperature measured by a sensor on an inlet port of the respective reference IDG;
a temperature measured by a sensor on an outlet port of the respective reference IDG;
a frequency of electrical signal produced by the respective reference IDG;
a temperature of oil supplying the jet engine with which the respective reference IDG is associated; and
combinations thereof;
carry out a training, in supervisor mode, of a machine learning artificial intelligence with the reference data, with a first classification run for training to detect potential malfunctions suffered by IDGs and a second classification run for training to determine causes of the malfunctions, where relevant;
use in production the machine learning artificial intelligence after validation of the training;
collect data to be analyzed, the data being the same as the reference data, relating to measurements made and recorded during flights of the aircraft comprising the IDG to be monitored; and
use the machine learning artificial intelligence that has been trained to predict a potential malfunction suffered by the IDG to be monitored and predict, where relevant, a cause of the malfunction with, respectively, the said first and second runs, using the data to be analyzed.

Assignees

• Airbus Operations Sas

Inventors

• Brouqui Manon
• Bradier Matthieu
• Benhamou Eric

Classifications

• G01R31/34

  Testing dynamo-electric machines · CPC title

• B64D2045/0085

  Devices for aircraft health monitoring, e.g. monitoring flutter or vibration · CPC title

• H02K11/20

  for measuring, monitoring, testing, protecting or switching (rectifiers H02K11/04; power electronics H02K11/33) · CPC title

• G06N3/08

  Learning methods · CPC title

• G05B23/0281

  Quantitative, e.g. mathematical distance; Clustering; Neural networks; Statistical analysis · CPC title