Predictive maintenance system using avionics ethernet networks

Patent metadata

Field	Value
Publication number	US-11465779-B2
Application number	US-201916540740-A
Country	US
Kind code	B2
Filing date	Aug 14, 2019
Priority date	Aug 14, 2019
Publication date	Oct 11, 2022
Grant date	Oct 11, 2022

Abstract

A predictive maintenance system is disclosed. The system includes a network of analog and digital sensors, each sensor configured for measuring telemetry data associated with temperature levels, voltage levels, current levels, and other analog or digital parameters. The system includes microprocessors for receiving the (digitized) analog and digital telemetry data, tabulating and timestamping the raw telemetry datasets. The microprocessors compress the raw data and reduce its dimensionality by generating principal component sets from the raw data based on scalar parameters corresponding to machine learning algorithms stored to memory, the principal component sets capturing a majority of variances within the raw data. The principal component sets are organized into data packets including identifiers for the relevant algorithms. The data packets are transmitted via real time networks for either onboard storage or ground-based analysis.

First claim

We claim:

1. A predictive maintenance system, comprising:
one or more sensors disposed within an aircraft, each sensor configured to 1 ) monitor one or more parameters associated with the aircraft and 2 ) collect at least one sample set of telemetry data associated with the parameters;
at least one memory configured for storing:
one or more machine learning algorithms;
at least one algorithm identifier corresponding to each machine learning algorithm; and
a plurality of scalar parameters corresponding to each machine learning algorithm; and

at least one microprocessor in communication with the memory and with the one or more sensors, the microprocessor configured to:
generate at least one set of tabulated telemetry data based on the at least one sample set of telemetry data;
generate at least one set of principal components corresponding to the sample set of tabular telemetry data according to at least one selected scalar parameter of the plurality of scalar parameters; and
generate at least one data payload comprising 1) the set of principal components and 2) the at least one algorithm identifier corresponding to the at least one selected scalar parameter.

2. The predictive maintenance system of claim 1 , wherein the one or more sensors include at least one digital sensor configured to collect digital telemetry data associated with one or more digital parameters.

3. The predictive maintenance system of claim 1 , wherein the one or more sensors include:
at least one analog sensor configured to collect analog telemetry data associated with one or more analog parameters; and
at least one analog/digital converter (ADC) in communication with the at least one analog sensor and with the at least one microprocessor, the ADC configured to digitize the analog telemetry data.

4. The predictive maintenance system of claim 3 , wherein the one or more analog parameters include at least one of a temperature, a voltage, and a current level.

5. The predictive maintenance system of claim 1 , wherein the at least one microprocessor is configured to generate the at least one set of tabulated telemetry data by:
timestamping each element of the at least one sample set of telemetry data; and
tabulating each timestamped element of the at least one sample set of telemetry data.

6. The predictive maintenance system of claim 1 , wherein:
the system is embodied within a line replaceable unit (LRU) of the aircraft, the LRU further comprising:
at least one internal memory in communication with the microprocessor, the internal memory configured for storage of one or more of the at least one set of principal components and the at least one data payload.

7. The predictive maintenance system of claim 1 , further comprising:
at least one network interface in communication with the microprocessor, the network interface configured to transmit the at least one data payload via a real-time network of the aircraft to at least one of 1) a data concentrator disposed aboard the aircraft and 2) a control facility external to the aircraft.

8. The predictive maintenance system of claim 1 , wherein the sample set of telemetry data is a first sample set, and wherein:
the one or more sensors are configured to collect at least one second sample set of telemetry data; and
the at least one microprocessor is configured to:
center the at least one second sample set of telemetry data;
determine at least one covariance matrix corresponding to the second sample set;
determine at least one eigenvalue of the covariance matrix, the eigenvalue corresponding to at least one eigenvector;
select one or more eigenvalues of the at least one determined eigenvalue based on a predetermined desired variance threshold, and
modify the at least one scalar parameter based on at least one eigenvector corresponding to the one or more selected eigenvalues.

9. A method for predictive maintenance, comprising:
collecting, via at least one aircraft sensor, at least one first sample set of telemetry data;
determining, via the at least one microprocessor, at least one covariance matrix associated with the first sample set of telemetry data;
determining, via the at least one microprocessor, at least one eigenvalue of the covariance matrix, the eigenvalue corresponding to at least one eigenvector;
selecting, via the at least one microprocessor, one or more eigenvalues of the at least one determined eigenvalue based on a predetermined desired variance threshold,
defining at least one scalar parameter based on at least one eigenvector corresponding to the one or more selected eigenvalues, the scalar parameter associated with a machine learning algorithm;
collecting at least one second sample set of telemetry data from the at least one aircraft sensor;
generating at least one set of tabulated telemetry data by tabulating the at least one second sample set of telemetry data via the at least one microprocessor;
generating one or more principal components corresponding to the at least one set of tabulated telemetry data via the at least one microprocessor according to the at least one scalar parameter;
generating one or more data payloads via the at least one microprocessor, each data packet comprising the one or more principal components and at least one identifier corresponding to the associated machine learning algorithm; and
transmitting the one or more data payloads via at least one real-time avionics network.

10. The method of claim 9 , wherein determining, via the at least one microprocessor, at least one covariance matrix associated with the second sample set of telemetry data includes:
centering, via the at least one microprocessor, the at least one second sample set of telemetry data; and
determining at least one covariance matrix associated with the centered telemetry data.

11. The method of claim 9 , wherein generating at least one set of tabulated telemetry data by tabulating the at least one second sample set of telemetry data via the at least one microprocessor includes:
timestamping the at least one second sample set of telemetry data via the at least one microprocessor; and
tabulating the at least one set of timestamped telemetry data via the at least one microprocessor.

12. The method of claim 9 , wherein collecting at least one second sample set of telemetry data from the at least one aircraft sensor includes:
collecting at least one sample set of analog telemetry data from at least one analog aircraft sensor; and
digitizing the at least one sample set of analog telemetry data via one or more analog-digital converters (ADC) in communication with the at least one analog sensor.

13. The method of claim 9 , wherein collecting at least one second sample set of telemetry data from the at least one aircraft sensor includes:
collecting at least one sample set of digital telemetry data from at least one digital aircraft sensor.

14. The method of claim 9 , wherein transmitting the one or more data payloads via at least one real-time avionics network includes:
transmitting the one or more data payloads to at least one of a data concentrator aboard the aircraft and a ground control facility external to the aircraft.

15. The method of claim 9 , wherein the at least one covariance matrix is a first covariance matrix, the eigenvalue is a first eigenvalue, and the eigenvector is a first eigenvector, further comprising:
collecting, via the at least one aircraft sensor, at least one third sample set of telemetry data;
determining, via the at least one microprocessor, at least one second covariance matrix associated with the third sa

Assignees

• Rockwell Collins Inc

Inventors

• Kaplan Daniel J
• Hohensee Christopher A
• Wijesekara Gamachchige Yeshani D
• Harvey Chadwick K J

Classifications

• G06N20/00

  Machine learning · CPC title

• H04Q9/02

  Automatically-operated arrangements · CPC title

• H04Q2209/10

  using a centralized architecture · CPC title

• H04L67/12

  specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks · CPC title

• G06F11/3013

  where the computing system is an embedded system, i.e. a combination of hardware and software dedicated to perform a certain function in mobile devices, printers, automotive or aircraft systems (testing or monitoring of control systems or parts thereof G05B23/02) · CPC title