System and method for analyzing engine test data in real time

Patent metadata

Field	Value
Publication number	US-11248989-B2
Application number	US-201916448365-A
Country	US
Kind code	B2
Filing date	Jun 21, 2019
Priority date	Jun 21, 2019
Publication date	Feb 15, 2022
Grant date	Feb 15, 2022

Abstract

A system for providing real time aircraft engine sensor analysis includes a computer system configured to receive an engine operation data set in real time. The computer system includes a machine learning based analysis tool and a user interface configured to display a real time analysis of the engine operation data set. The user interface includes at least one portion configured to identify a plurality of anomalies in the engine operation data set.

First claim

The invention claimed is:

1. A system for providing real time aircraft engine sensor analysis comprising:
a computer system configured to receive a plurality of engine operation data sets in real time, the computer system including a machine learning based analysis tool;
a user interface including a main screen, the main screen including an operation tile corresponding to each of the engine operation data sets in the plurality of engine operation data sets, the user interface being configured to display a real time analysis of the engine operation data set, and the user interface including at least one portion configured to identify a plurality of anomalies in the engine operation data set.

2. A system for providing real time aircraft engine sensor analysis comprising:
a computer system configured to receive an engine operation data set in real time, the computer system including a machine learning based analysis tool;
a user interface configured to display a real time analysis of the engine operation data set, the user interface including at least one portion configured to identify a plurality of anomalies in the engine operation data set;
wherein the computer system is configured to receive a plurality of engine operation data sets in real time, and wherein the user interface includes a main screen, the main screen including an operation tile corresponding to each of the engine operation data sets in the plurality of engine operation data sets; and
wherein each operation tile includes a display indicative of an engine operation completion percentage and at least one text information display.

3. A system for providing real time aircraft engine sensor analysis comprising:
a computer system configured to receive an engine operation data set in real time, the computer system including a machine learning based analysis tool;
a user interface configured to display a real time analysis of the engine operation data set, the user interface including at least one portion configured to identify a plurality of anomalies in the engine operation data set;
wherein the computer system is configured to receive a plurality of engine operation data sets in real time, and wherein the user interface includes a main screen, the main screen including an operation tile corresponding to each of the engine operation data sets in the plurality of engine operation data sets; and
wherein each operation tile is expandable, and wherein an expanded engine operation tile includes an engine operation completion percentage and a sensor map.

4. The system of claim 3 , wherein the sensor map is configured to highlight sensors corresponding to anomalous sensor readings in real time by illuminating a position on the sensor map corresponding to a position of the sensor.

5. The system of claim 4 , wherein the sensor map comprises a plurality of engine cross sections along an engine centerline axis.

6. The system of claim 4 , wherein a prominence of the illumination corresponds to a time since the occurrence of the anomalous sensor reading.

7. A system for providing real time aircraft engine sensor analysis comprising:
a computer system configured to receive an engine operation data set in real time, the computer system including a machine learning based analysis tool wherein the machine learning based analysis tool is a multi-input multi output deep auto-encoder (MIMODAE) tool; and
a user interface configured to display a real time analysis of the engine operation data set, the user interface including at least one portion configured to identify a plurality of anomalies in the engine operation data set.

8. The system of claim 7 , wherein the MIMODAE tool includes an input, and wherein the input is configured to be expanded across a decoder portion.

9. The system of claim 8 , wherein the decoder portion includes at least three decoder layers.

10. The system of claim 7 , wherein the machine learning based analysis tool is trained by providing a plurality of normalized data sets to an encoder portion of the MIMODAE tool.

11. The system of claim 10 , wherein the normalized data sets are outputs of at least one engine sensor and wherein the machine learning based analysis tool learns at least one of a correlation between the outputs of at least two engine sensors at a single point in time or a correlation between the output of a single engine sensor at at least two different points in time.

12. The system of claim 10 , wherein the encoder portion includes at least three encoder layers.

13. The system of claim 12 , wherein each encoder layer in the at least three encoder layers is configured to determine relationships between data points within the plurality of normalized data sets at a given time, and wherein the relationships are indicative of a presence of one or more anomalies.

14. A method for analyzing engine operation data in real time comprising:
receiving at least one engine operations data set in real time, wherein at least one engine operations data set includes at least 1000 sensor data points at each time entry;
analyzing the at least one engine operations data set using a multi-input multi output deep auto-encoder (MIMODAE) tool, wherein the MIMODAE includes at least three encoder layers configured to learn relationships among sensor outputs in a plurality of nominal test data sets, and at least three decoder layers configured to analyze relationships between data points in the at least one engine operations data set and determine a presence of an anomaly when a relationship between data points in the at least one engine operations data set diverges from an expected relationship based on a learned relationship among the sensor outputs in the plurality of nominal test data sets.

15. The method of claim 14 , wherein the at least one engine operations data set includes a plurality of engine operations data sets, and wherein the plurality of engine operations data sets are simultaneously analyzed.

16. The method of claim 14 , further comprising notifying a user of the anomaly in response to determining the presence of the anomaly.

17. The method of claim 16 , wherein notifying the user comprises changing a display of a user interface.

18. The method of claim 17 , wherein changing the display comprises illuminating a sensor position on a sensor map, and wherein the sensor position corresponds to the sensor providing the anomaly.

19. The method of claim 17 , wherein changing the display comprises altering an appearance of an operations tile corresponding to an engine operation data set in the at least one engine operations data set including the anomaly.

Assignees

• United Technologies Corp
• Raytheon Tech Corp

Inventors

• Somanath Nagendra
• Urban Justin R
• Giering Michael J
• Long Quan
• Dorgan Alexandria

Classifications

• G06N3/045

  Combinations of networks · CPC title

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• G06N3/0455

  Auto-encoder networks; Encoder-decoder networks · CPC title

• G05B23/024 Primary

  Quantitative history assessment, e.g. mathematical relationships between available data; Functions therefor; Principal component analysis [PCA]; Partial least square [PLS]; Statistical classifiers, e.g. Bayesian networks, linear regression or correlation analysis; Neural networks · CPC title

• G06N3/088

  Non-supervised learning, e.g. competitive learning · CPC title