Methods and systems for diagnosing machinery

• US 20060241910A1
• Filed: 04/21/2005
• Published: 10/26/2006
• Est. Priority Date: 04/21/2005
• Status: Active Grant

• Current Assignee: General Electric Company
• Original Assignee: General Electric Company
• Inventors: Grant, John Wesley, Malakhova, Olga, Littrell, Nathan Bowman, Petchenev, Alexei
• Granted Patent: US 7,222,048 B2
• Time in Patent Office: Days
• US Class Current: 702/184
• CPC Class Codes: G05B 23/0232 based on qualitative trend analysis, e.g. system evolution
• CPC Class Codes: G05B 23/0283 Predictive maintenance, e.g. involving the monitoring of a system and, based on the monitoring results, taking decisions on the maintenance schedule of the monitored system; Estimating remaining useful life [RUL] preventive maintenance, i.e. planning maintenance according to the available resources without monitoring the system G06Q10/06

Abstract

A method of predicting the remaining operational life of a component including generating a plurality of component data utilizing a machinery monitoring system including a database having at least one rule set, the component data indicative of a least one operational parameter of the component, trending the component data utilizing the rule set, and extrapolating the trended data to facilitate predicting the remaining operational life of the component.

Claims

1. A method of predicting the remaining operational life of a component, said method comprising:

generating a plurality of component data samples utilizing a machinery monitoring system including a database having at least one rule set, the component data samples indicative of at least one operational parameter of the component;

communicating the component data samples to a first auto-buffer;

trending the component data samples utilizing the rule set;

extrapolating the trended data to facilitate predicting the remaining operational life of the component; and

outputting the remaining operational life of the component.

2. A method in accordance with claim 1 further comprising generating a time when the extrapolated data will exceed a predetermined threshold.

3. A method in accordance with claim 1 further comprising:

utilizing a data collecting rule to generate the component data samples;

utilizing the first auto-buffer to generate a plurality of data points; and

communicating the plurality of data points to a first manual buffer.

4. A method in accordance with claim 3 wherein utilizing the first auto-buffer to generate a plurality of data points further comprises:

determining an average of the first data points utilizing the first auto-buffer; and

communicating the average to the first manual buffer; and

utilizing the average within the first manual buffer to generate a graph illustrating the prediction.

5. A method in accordance with claim 3 further comprising:

verifying that the data samples are within a predefined range;

deleting the data samples if the data samples are not within the predefined range; and

generating an indication that the prediction is not accurate if at least one of the data samples is not within the predefined range.

6. A method in accordance with claim 5 further comprising generating an indication that the prediction is invalid after a predetermined length of time.

7. A method in accordance with claim 1 further comprising:

collecting a data sample every minute utilizing a first auto-buffer;

averaging sixty data samples to generate a single data point utilizing the first auto-buffer; and

extrapolating at least eight data points; and

buffering the eight data points to facilitate generating a relatively smooth graphical illustration predicting the remaining operational life of the component.

8. A method in accordance with claim 7 further comprising extrapolating the eight data points using a least squares fit method to facilitate generating a curve representative of the extrapolate data points.

9. A method in accordance with claim 8 further comprising generating at least one of a polynomial curve, a exponential curve, and a Weibull curve using the eight data points.

10. A method in accordance with claim 2 wherein generating a time when the extrapolated data will exceed a predetermined threshold further comprises generating a time in accordance with:

11. A computer program encoded on a computer readable medium for predicting the remaining operational life of a component, said computer program configured to:

generate a plurality of component data samples utilizing a machinery monitoring system including a database having at least one rule set, the component data samples indicative of at least one operational parameter of the component;

communicate the component data samples to a first auto-buffer;

trend the component data samples utilizing the rule set;

extrapolate the trended data to facilitate predicting the remaining operational life of the component; and

output the remaining operational life of the component.

12. A computer program in accordance with claim 11 further configured to generate a time when the extrapolated data will exceed a predetermined threshold.

13. A computer program in accordance with claim 11 further configured to:

utilize a data collecting rule to generate the component data samples;

utilize the first auto-buffer to generate a plurality of data points; and

communicate the plurality of data points to a first manual buffer.

14. A computer program in accordance with claim 13 further configured to:

determine an average of the first data points utilizing the first auto-buffer; and

communicate the average to the first manual buffer; and

utilize the average within the first manual buffer to generate a graph illustrating the prediction.

15. A computer program in accordance with claim 13 further configured to:

verify that the data samples are within a predefined range;

delete the data samples if the data samples are not within the predefined range; and

generate an indication that the prediction is not accurate if at least one of the data samples is not within the predefined range.

16. A computer program in accordance with claim 15 further configured to generate an indication that the prediction is invalid after a predetermined length of time.

17. A computer program in accordance with claim 11 further configured to:

collect a data sample every minute utilizing a first auto-buffer;

average sixty data samples to generate a single data point utilizing the first auto-buffer; and

extrapolate at least eight data points; and

buffer the eight data points to facilitate generating a relatively smooth graphical illustration predicting the remaining operational life of the component.

18. A computer program in accordance with claim 17 further configured to extrapolate the eight data points using a least squares fit method to facilitate generating a curve representative of the extrapolate data points.

19. A computer program in accordance with claim 18 further configured to generate at least one of a polynomial curve, a exponential curve, and a Weibull curve using the eight data points.

20. A computer program in accordance with claim 12 further configured to generate a time when the extrapolated data will exceed a predetermined threshold in accordance with: