Method and system for determining motor reliability

• US 6,834,256 B2
• Filed: 08/30/2002
• Issued: 12/21/2004
• Est. Priority Date: 08/30/2002
• Status: Active Grant

• Current Assignee: General Electric Company
• Original Assignee: General Electric Company
• Inventors: Repoff, Thomas Paul, Sutherland, Hunt Adams, House, Michael Brynn, Flickinger, Gregory Lee, Fritz, Calvin Thomas
• Primary Examiner(s): WACHSMAN, HAL D
• Application Number:
  US10/231,084
• Publication Number:
  US 20040044499A1
• Time in Patent Office: 844 Days
• Field of Search: 345/713, 703/2-4, 703/13, 702/181, 702/34-36, 702/56, 702/58, 702/59, 702/64, 702/65, 702/113, 702/122, 702/179, 702/182-185, 702/187-188, 324/772, 324/546, 324/545, 324/522, 324/525, 324/551, 388/909, 318/565, 318/434, 318/806, 340/500, 340/514, 340/635, 340/636.13, 340/648, 340/657, 340/664, 340/680, 340/682, 340/683, 340/3.1, 361/23, 361/24, 706/14, 706/50, 706/911, 706/912, 706/914, 706/915, 700/79, 700/21, 700/293, 700/294
• US Class Current: 702/181
• CPC Class Codes: G01R 31/343 in operation
• CPC Class Codes: H02P 29/02 Providing protection against overload without automatic interruption of supply protection against faults of stepper motors H02P8/36

Abstract

The reliability and remaining time before failure for electric motor systems and the like is determined with an ascertained statistical confidence based on continually monitored system parameters and a database comprised of historical failure mode/cause and repair data concerning the same motor and/or similar motor systems. A computational process is followed that relies upon hierarchical failure-tree structures, Weibull probability distributions and directed causal networks and which updates the probabilities and models used to forecast the remaining time before system failure based on the monitored parameters. Motor system operational parameter data is continually monitored at the site of the motor system and up-loaded via wireless or other communications network to a remote computer system where a database is maintained and reliability analysis is performed.

Claims

1. A method for determining reliability and a remaining time before failure with statistical confidence for an electric motor system, comprising steps performed by a computer system of:

acquiring and organizing historical motor repair data corresponding to the motor system into a computer system database;

continually monitoring and uploading operational parameter data from sensors at the motor system to the computer system;

performing failure analysis based on a composite of reliability probability distributions corresponding to predetermined sub-populations of historical motor system failure causes;

developing a causal network for modeling reliability of one or more motor system components and assessing motor system component condition based on the causal network;

performing an integrated causal network and reliability analysis of the motor system, wherein results from performing failure analysis based on historical data are integrated with analysis results based on the causal network to compute a quantitative value for a time remaining before failure with an ascertained statistical confidence; and

outputting the computed value on a display device associated with the computer system.

2. The method according to claim 1 wherein a Bayesian Belief Network is used as a causal network for modeling reliability.

3. The method according to claim 1 wherein a Petrinet is used as a causal network for modeling reliability.

4. The method according to claim 1 wherein the step of performing failure analysis comprises modeling a failure rate of one or more system components using a Weibull probability distribution function.

5. The method according to claim 1 wherein a portable computer is used to collect motor operational parameter data from sensors at the motor system site and collected data is uploaded via landline or a wireless communications link to a remote computer system.

6. The method according to claim 1 further including the step of providing a graphical user interface that displays remaining time before failure and/or predicted motor life as a graphical odometer.

7. The method according to claim 1 further including the step of providing a graphical user interface that displays remaining time before failure and/or predicted motor life as a moving bar-graph.

8. A method for assessing reliability and determining, with a statistical degree of confidence, a quantitative value for time remaining before failure for an electric motor system and/or electric motor driven equipment, comprising the steps of:

storing motor system component failure data in a computer data base, said failure rate data based on historical test data and/or failure analysis of similar or identical motor systems and motor components;

performing a motor system fault-tree analysis based on data assembled in the component failure data base;

developing one or more motor system component fault models based on said motor system fault tree analysis;

modeling the historical failure data of one or more motor system components using one or more conventional probability distribution functions;

developing a directed casual inference model and selecting system variables used for collecting data based on said casual inference model;

developing a conditional probability matrix comprising failure probabilities corresponding to possible failure events for said motor system;

developing an overall motor system reliability model based on said casual inference model and said conditional probability matrix; and

computing a statistical assessment of reliability of the motor system and a quantitative value for time remaining before failure using said overall motor system reliability model.

9. The method according to claim 8 wherein, as a result of determining said statistical assessment of reliability, one or more motor system components are identified as a most probable candidate for needing service or replacement at a predicted time of system failure.

10. The method according to claim 8 wherein the motor system failure data is stored on a computer system remote from the site of the motor system and sensor data is forwarded via wireline or wireless communications link to the remote computer system.

11. The method according to claim 8 wherein the historical failure data is modeled using one or more Weibull probability distribution functions and further includes the step of determining appropriate values for individual Weibull shape parameters to fit said Weibull probability distribution functions to failure data corresponding to individual system components.

12. The method according to claim 8 further comprising the step of updating the historical failure information data base with information obtained through root cause failure analysis performed at times of preventive maintenance or repair of the motor system.

13. The method according to claim 8 wherein motor system operational parameter data is acquired using one or more operational parameter sensors located at the motor system site.

14. The method according to claim 8 wherein a portable computer is used to collect motor operational parameter data from sensors at the motor system site and collected data is uploaded via wireline or a wireless communications link to a remote computer system.

15. The method according to claim 8 further including the step of providing a graphical user interface that displays remaining and/or predicted motor life as a graphical odometer.

16. The method according to claim 8 further including the step of providing a graphical user interface that displays remaining and/or predicted motor life as a moving bar-graph.

17. The method according to claim 8 wherein the historical failure data of one or more system components is modeled using a Weibull probability distribution function.

18. The method according to claim 8 wherein a directed casual inference model is developed using a Bayesian Belief Network.

19. The method according to claim 8 wherein a directed casual inference model is developed using a Petrinet casual network for modeling reliability.

20. A computer-implemented method for determining electrical motor system reliability with an ascertained statistical confidence, comprising:

modeling historical failure data of one or more system components using one or more conventional probability distribution functions;

developing a motor system reliability model based on historical failure data, causal inference models of motor system failure and conditional probabilities describing possible failure events;

collecting motor operational parameter data acquired via motor system sensors located at the motor system site;

evaluating a present state of motor system reliability using the motor system reliability model and collected operational parameter data; and

computing a statistical assessment of motor system reliability and a quantitative value for time remaining before failure using the motor system reliability model.

21. The method of claim 20 wherein the system reliability model is based at least in part on a motor system component reliability hierarchy, wherein the motor system component reliability hierarchy is determined from a motor system fault tree.

22. The method of claim 20 wherein a quantitative value for a time remaining before motor system failure is computed and a statistical reliability of the motor system is provided within an ascertained degree of statistical confidence.

23. The method of claim 20 wherein failure data is based at least in part on historical test data and/or failure analysis of similar or identical motor systems and motor components.

24. The method of claim 20 wherein historical failure data of one or more system components is modeled using a Weilbull probability distribution function.

25. The method of claim 20 wherein a Bayesian Belief Network is used to construct a casual inference model for failure events.

26. The method of claim 20 wherein a Petrinet casual network is used to construct a casual inference model for failure events.

27. The method according to claim 20 wherein a portable computer is used to collect motor operational parameter data from sensors at the motor system site and collected data is uploaded via landline or a wireless communications link to a remote computer system.

28. The method according to claim 20 further including the step of providing a graphical user interface that displays remaining time before failure and/or predicted motor life as a graphical odometer.

29. The method according to claim 20 further including the step of providing a graphical user interface that displays remaining time before failure and/or predicted motor life as a moving bar-graph.