Dynamic data filtering system and method

• US 8,712,929 B1
• Filed: 03/29/2010
• Issued: 04/29/2014
• Est. Priority Date: 11/30/2007
• Status: Active Grant

• Current Assignee: Intellectual Assets LLC
• Original Assignee: Intellectual Assets LLC
• Inventors: Bickford, Randall L., Palnitkar, Rahul M.
• Primary Examiner(s): Gaffin, Jeffrey A
• Assistant Examiner(s): Sugent, James F
• Application Number:
  US12/798,128
• Time in Patent Office: 1,492 Days
• Field of Search: 706/12, 702/104, 700/286
• US Class Current: 706/12
• CPC Class Codes: G01D 3/022 having an ideal characteristic, map or correction data stored in a digital memory
• CPC Class Codes: G05B 23/0235 based on a comparison with predetermined threshold or range, e.g. "classical methods", carried out during normal operation; threshold adaptation or choice; when or how to compare with the threshold
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06Q 50/06 Energy or water supply

Abstract

A computer-implemented dynamic data filtering system and method for selectively choosing operating data of a monitored asset that modifies or expands a learned scope of an empirical model of normal operation of the monitored asset while simultaneously rejecting operating data of the monitored asset that is indicative of excessive degradation or impending failure of the monitored asset, and utilizing the selectively chosen data for adaptively recalibrating the empirical model to more accurately monitor asset aging changes or operating condition changes of the monitored asset.

Claims

1. A computer-implemented dynamic data filtering method, said method comprising the steps of:

calibrating at least one data filter having at least one statistically based test for defining at least one predefined criterion of good data quality wherein the calibrating includes utilizing initial asset operating data values to define an initial scope of operation of an asset;

acquiring additional asset operating data values for an additional operation of the asset; and

filtering the acquired additional asset operating data values using the at least one data filter utilizing the at least one statistically based test to define the at least one predefined criterion of good data quality for selectively choosing asset operating data values that meet the at least one predefined criterion of good data quality;

segregating the selectively chosen asset operating data values that meet the at least one predefined criterion of good data quality; and

storing the segregated selectively chosen asset operating data values that meet the at least one predefined criterion of good data quality for subsequent use in adjusting the learned scope of at least one previously trained model.

2. The computer-implemented method of claim 1 wherein filtering the acquired additional asset operating data values utilizing the at least one data filter includes utilizing a probability estimation method using sequential discounting expectation maximization for obtaining a score for at least one of the acquired additional asset operating data values wherein the obtained scores define the at least one redefined criterion of data quality.

3. The computer-implemented method of claim 1 wherein filtering the acquired additional asset operating data values utilizing the at least one data filter includes utilizing an adaptive sequential probability hypotheses test method to define the at least one predefined criterion of data quality.

4. The computer-implemented method of claim 1 further comprising a step of iteratively repeating the acquiring, filtering, segregating, and storing steps as a function of a user-specified period or on user demand.

5. A non-transitory computer-readable medium containing computer-executable instructions that, when executed by a processor, cause the processor to perform a dynamic data filtering method, said method comprising:

calibrating at least one data filter having at least one statistically based test for defining at least one predefined criterion of good data quality wherein the calibrating includes utilizing initial asset operating data values to define an initial scope of operation of an asset;

acquiring additional asset operating data values for an additional operation of the asset; and

filtering the acquired additional asset operating data values using the at least one data filter utilizing the at least one statistically based test to define the at least one predefined criterion of good data quality for selectively choosing asset operating data values that meet the at least one predefined criterion of good data quality;

segregating the selectively chosen asset operating data values that meet the at least one predefined criterion of good data quality; and

storing the segregated selectively chosen asset operating data values that meet the at least one predefined criterion of good data quality for subsequent use in adjusting the learned scope of at least one previously trained model.

6. The non-transitory computer-readable medium of claim 5 wherein filtering the acquired additional asset operating data values utilizing the at least one data filter includes utilizing a probability estimation method using sequential discounting expectation maximization for obtaining a score for at least one of the acquired additional asset operating data values wherein the obtained scores define the at least one predefined criterion of data quality.

7. The non-transitory computer-readable medium of claim 5 wherein filtering the acquired additional asset operating data values utilizing the at least one data filter includes utilizing an adaptive sequential probability hypotheses test method to define the at least one predefined criterion of data quality.

8. The non-transitory computer-readable medium of claim 5 further comprising a step of iteratively repeating the acquiring, filtering, segregating, and storing steps as a function of a user-specified period or on user demand.

9. A dynamic data filtering system, said system comprising:

means for calibrating at least one data filter having at least one statistically based test for defining at least one predefined criterion of good data quality wherein said calibrating means comprises means for utilizing initial asset operating data values to define an initial scope of operation of an asset;

means for acquiring additional asset operating data values for an additional operation of the asset; and

means for filtering said acquired additional asset operating data values using the at least one data filter wherein said filtering means comprises means for utilizing said at least one statistically based test to define said at least one predefined criterion of good data quality for selectively choosing asset operating data values that meet said at least one predefined criterion of good data quality;

means for segregating said selectively chosen asset operating data values that meet said at least one predefined criterion of good data quality; and

means for storing said segregated selectively chosen asset operating data values that meet said at least one predefined criterion of good data quality for subsequent use in adjusting the learned scope of at least one previously trained model.

10. The system of claim 9 wherein said means for filtering said acquired additional asset operating data values utilizing said at least one data filter includes means for utilizing at least one statistically based test to define said at least one predefined criterion of good data quality.

11. The system of claim 9 wherein said means for filtering said acquired additional asset operating data values utilizing the at least one data filter includes means for utilizing a probability estimation method using sequential discounting expectation maximization for obtaining a score for at least one of said acquired additional asset operating data values wherein said obtained scores define said at least one predefined criterion of data quality.

12. The system of claim 9 wherein said means for filtering said acquired additional asset operating data values utilizing the at least one data filter includes means for utilizing an adaptive sequential probability hypotheses test method to define said at least one predefined criterion of data quality.

13. A computer-implemented dynamic data filtering method, said method comprising the steps of:

calibrating at least one data filter having at least one statistically based test utilizing initial asset operating data values wherein the initial asset operating data values are used for defining an initial scope of operation of an asset;

acquiring additional asset operating data values for an additional operation of the asset;

using the at least one data filter having the at least one statistically based test defining at least one predefined criterion of good data quality for assigning a measure of data quality to at least one of the acquired additional asset operating data values;

segregating at least one of the additional asset operating data values based on the measure of data quality assigned to the at least one additional asset operating data values; and

storing the segregated at least one additional asset operating data values for defining an additional scope of operation of the asset.

14. The computer-implemented dynamic data filtering method of claim 13 wherein the step of using the at least one data filter having the at least one statistically based test defining at least one predefined criterion of good data quality for assigning a measure of data quality to at least one of the acquired additional asset operating data values includes the steps of:

comparing at least one of the acquired additional asset operating data values to at least one value computed based on at least one empirical model; and

assigning a measure of data quality to at least one of the acquired additional asset operating data values based on the comparing step.

15. The computer-implemented dynamic data filtering method of claim 13 further including the step of utilizing the stored segregated at least one additional asset operating data values for recalibrating at least one data filter.

16. The computer-implemented dynamic data filtering method of claim 13 further including the step of utilizing the stored segregated at least one additional asset operating data values for recalibrating at least one empirical model.

17. The computer-implemented dynamic data filtering method of claim 13 further including the step of utilizing the stored segregated at least one additional asset operating data values for recalibrating at least one fault detector.