Health Management Using Distances for Segmented Time Series

Patent metadata

Field	Value
Publication number	US-2017193372-A1
Application number	US-201614989204-A
Country	US
Kind code	A1
Filing date	Jan 6, 2016
Priority date	Jan 6, 2016
Publication date	Jul 6, 2017
Grant date	—

Abstract

A method, apparatus, and system for evaluating health of a component of a vehicle. Time series data generated during operation of the vehicle is received by a computer system. The computer system transforms the time series data into a plurality of segments based on a selected state for the vehicle that is of interest. The computer system builds a prognostic distance matrix based on pairings formed using the plurality of segments. The prognostic distance matrix comprises distances that measure deviation of the component from nominal performance for the selected state. The computer system generates a digital prognosis for the component of the vehicle based on the prognostic distance matrix. The digital prognosis predicts whether a maintenance operation should be performed with respect to the component. The distances may be used to diagnose what maintenance operation should be performed.

First claim

What is claimed is:

1 . A method for evaluating a component of a vehicle, the method comprising:
receiving, by a computer system, time series data generated during operation of the vehicle; transforming, by the computer system, the time series data into a plurality of segments based on a selected state for the vehicle that is of interest; building, by the computer system, a prognostic distance matrix based on pairings formed using the plurality of segments,
wherein the prognostic distance matrix comprises distances that measure deviation of the component from nominal performance for the selected state; and
generating, by the computer system, a digital prognosis for the component of the vehicle based on the prognostic distance matrix, wherein the digital prognosis predicts whether a maintenance operation should be performed with respect to the component.

2 . The method of claim 1 further comprising:
generating, by the computer system, a diagnosis output based on at least a portion of the distances in the prognostic distance matrix, wherein the diagnosis output helps a user determine what maintenance operation should be performed.

3 . The method of claim 1 further comprising:
performing the maintenance operation for the component of the vehicle, wherein the maintenance operation comprises at least one of inspection, preventative maintenance, corrective maintenance, adaptive maintenance, in-use maintenance, repair, rework replacement, or overhaul.

4 . The method of claim 1 , wherein transforming, by the computer system, the plurality of segments into the prognostic distance matrix comprises:
building a plurality of pairings using the plurality of segments; and computing a distance for each pairing of the plurality of pairings.

5 . The method of claim 4 , wherein building the plurality of pairings comprises:
building a segment-segment pairing between a first segment extracted from a portion of the time series data generated by a first sensor during the selected state and a second segment extracted from another portion of the time series data generated by a second sensor during the selected state.

6 . The method of claim 4 , wherein building the plurality of pairings comprises:
identifying a nominal segment for the selected state; and building a segment-nominal pairing between a segment extracted from a portion of the time series data generated during the selected state and a nominal segment.

7 . The method of claim 6 , wherein identifying the nominal segment comprises:
estimating expected values for the selected state to form the nominal segment.

8 . The method of claim 4 , wherein building the plurality of pairings comprises:
identifying a shapelet for use in analyzing a segment of the plurality of segments, wherein the shapelet is a subsequence that is present in either non-nominal performance or nominal performance but not both; and building a shapelet-segment pairing between the segment and the shapelet.

9 . The method of claim 8 , wherein computing the distance comprises:
computing the distance for the shapelet-segment pairing by identifying a best match between the shapelet and one of a plurality of sections of the segment.

10 . The method of claim 1 , wherein building, by the computer system, the prognostic distance matrix comprises:
building the prognostic distance matrix for segment-segment pairings, wherein the distances included in the prognostic distance matrix are computed for each distance type of a number of different distance types and wherein the distances for the each distance type are computed for each selected state of a number of selected states.

11 . The method of claim 1 , wherein building, by the computer system, the prognostic distance matrix comprises:
building the prognostic distance matrix for segment-nominal pairings, wherein the distances included in the prognostic distance matrix are computed for a number of different parameters for each selected state in a number of selected states.

12 . The method of claim 1 , wherein building, by the computer system, the prognostic distance matrix comprises:
building the prognostic distance matrix for segment-nominal pairings, wherein the distances included in the prognostic distance matrix are computed for a number of different types of shapelets for each selected state of a number of selected states and wherein the distances computed for the number of selected states are computed for each parameter of a number of different parameters.

13 . The method of claim 1 , wherein generating, by the computer system, the digital prognosis comprises:
analyzing, by a set of machine learning algorithms, the prognostic distance matrix; and generating a number of prognostic indicators based on the prognostic distance matrix that are included in the digital prognosis.

14 . The method of claim 1 , wherein generating, by the computer system, the digital prognosis comprises:
analyzing, by a set of machine learning algorithms, the prognostic distance matrix; and generating a final score for the component based on the prognostic distance matrix that is used to classify the component as either healthy or not healthy.

15 . The method of claim 1 further comprising:
displaying the digital prognosis in a graphical user interface on a display system, wherein the digital prognosis includes a plot of graphical indicators and wherein each of the graphical indicators represents a prognostic indicator with respect to a health of the component.

16 . The method of claim 1 , wherein transforming, by the computer system, the time series data into the plurality of segments comprises:
identifying each portion of data in the time series data that includes measurements generated during the selected state; and forming at least one segment using each portion of the data.

17 . The method of claim 1 further comprising:
selecting the selected state for the vehicle that is of interest, wherein the vehicle is an aircraft and wherein the selected state is selected from one of an entire flight of the aircraft, a particular phase of flight, a selected portion of a flight, a selected time period during flight, or a selected time period after a number of commands are sent from an aircraft control system to the component.

18 . The method of claim 1 further comprising:
displaying a visual representation of the time series data in a graphical user interface on a display system; and
displaying a number of distance features in association with the visual representation of the time series data that correspond to extreme distances.

19 . An apparatus comprising:
a segment generator module implemented in a computer system that receives time series data generated during operation of a vehicle and transforms the time series data into a plurality of segments based on a selected state for the vehicle that is of interest; a distance generator module implemented in the computer system that builds a prognostic distance matrix based on pairings formed using the plurality of segments,
wherein the prognostic distance matrix comprises distances that measure deviation of a component from nominal performance for the selected state; and
a health manager module implemented in the computer system that generates a digital prognosis for the component of the vehicle based on the prognostic distance matrix, wherein the digital prognosis predicts whether a maintenance operation should be performed with respect to the component and wherein the distances are

Assignees

• Boeing Co

Inventors

• Schimert James

Classifications

• G06N5/04 Primary

  Inference or reasoning models · CPC title

• G06Q10/20

  Administration of product repair or maintenance · CPC title

• G06N99/005

  Physics · mapped topic

• G06N20/00

  Machine learning · CPC title