Vibration analysis system and method for a machine

• US 20060122809A1
• Filed: 08/04/2005
• Published: 06/08/2006
• Est. Priority Date: 12/06/2004
• Status: Active Grant

• Current Assignee: Caterpillar Incorporated
• Original Assignee: Caterpillar Incorporated
• Inventors: Clarke, Burton R., Allgaier, Ryan P.
• Granted Patent: US 7,194,383 B2
• Time in Patent Office: Days
• US Class Current: 702/185
• CPC Class Codes: G01H 1/00 Measuring characteristics of vibrations in solids by using direct conduction to the detector G01H9/00, G01H11/00 take precedence
• CPC Class Codes: G01M 15/12 by monitoring vibrations

Abstract

A system and method for detecting and analyzing anomalies in a machine during operation. The system and method includes at least one sensor configured to detect characteristics of the machine indicative of machine vibration, at least one other sensor configured to detect characteristics of the machine indicative of other than machine vibration, a plurality of neural networks to receive input data, at least one neural network receiving vibration data from the at least one sensor, and at least one other neural network receiving non-vibration data from the at least one other sensor, and an expert system to receive output data from the neural networks and responsively analyze machine operation for anomalies.

Claims

1. A system for detecting and analyzing anomalies in a machine during operation, comprising:

at least one sensor configured to detect characteristics of the machine indicative of machine vibration;

at least one other sensor configured to detect characteristics of the machine indicative of other than machine vibration;

a plurality of neural networks to receive input data, at least one neural network receiving vibration data from the at least one sensor, and at least one other neural network receiving non-vibration data from the at least one other sensor, the at least one other neural network for comparing an influence of the non-vibration data on the vibration data; and

an expert system to receive output data from the neural networks and responsively analyze machine operation for anomalies.

2. A system, as set forth in claim 1, wherein at least one of the plurality of neural networks includes a weighting factor.

3. A system, as set forth in claim 2, wherein the weighting factor is an adjustable weighting factor.

4. A system, as set forth in claim 1, further including an autoregression algorithm to receive input data in time domain and convert the input data to frequency domain.

5. A system, as set forth in claim 4, wherein at least one neural network receives the converted data.

6. A system, as set forth in claim 1, further including at least one Fast Fourier Transform to receive input data and convert the input data from time domain to frequency domain.

7. A system, as set forth in claim 6, wherein at least one neural network receives the converted data.

8. A system, as set forth in claim 6, further including a plurality of band pass filters to split the converted data into frequency packets of data.

9. A system, as set forth in claim 8, further including a corresponding plurality of processors to receive each of the frequency packets and responsively determine a value indicative of a feature of the packet of data.

10. A system, as set forth in claim 8, wherein at least one neural network receives the frequency packets of data.

11. A system, as set forth in claim 9, wherein at least one neural network receives the determined values.

12. A system, as set forth in claim 9, wherein the at least one other neural network receiving non-vibration data receives at least one of the frequency packets of data and the determined values.

13. A system, as set forth in claim 1, further including at least one virtual channel to deliver at least a portion of the input data to a corresponding one of at least one algorithm.

14. A system, as set forth in claim 1, further including a multiple channel selector to select desired channels of data from the input data, process the selected channels, and deliver the selected channels to a correlation algorithm for analysis of interaction between the channels.

15. A system, as set forth in claim 14, wherein at least one neural network receives at least one processed channel.

16. A method for detecting and analyzing anomalies in a machine during operation, comprising the steps of:

receiving input data from a plurality of sensors, at least one sensor configured to detect characteristics of the machine indicative of machine vibration, and at least one other sensor configured to detect characteristics of the machine indicative of other than machine vibration;

delivering the input data to a plurality of neural networks, at least one neural network receiving vibration data, and at least one other neural network receiving non-vibration data, the at least one other neural network for comparing an influence of the non-vibration data on the vibration data;

delivering output data from the plurality of neural networks to an expert system; and

analyzing the output data for anomalies associated with machine operation.

17. A method, as set forth in claim 16, wherein receiving input data from a plurality of sensors includes the step of converting at least a portion of the input data from analog data to digital data.

18. A method, as set forth in claim 16, wherein receiving input data from a plurality of sensors includes the steps of:

conditioning at least a portion of the input data for proper voltage and current levels;

filtering the conditioned data to provide anti-aliasing; and

converting the filtered data from analog data to digital data.

19. A method, as set forth in claim 16, wherein delivering the input data to a plurality of neural networks includes the step of transmitting the input data to a hub containing a plurality of neural networks.

20. A method, as set forth in claim 16, wherein delivering output data from the plurality of neural networks includes the step of assigning a weighting factor to data from at least one neural network.

21. A method, as set forth in claim 16, wherein delivering the input data to a plurality of neural networks includes the steps of:

converting at least a portion of the input data from a time domain to a frequency domain; and

delivering the frequency domain data to at least one neural network.

22. A method, as set forth in claim 21, wherein converting at least a portion of the input data from a time domain to a frequency domain includes the step of applying an autoregression algorithm to the input data.

23. A method, as set forth in claim 21, wherein converting at least a portion of the input data from a time domain to a frequency domain includes the step of applying a Fast Fourier Transform to the input data.

24. A method, as set forth in claim 23, wherein delivering the input data to a plurality of neural networks includes the steps of:

splitting the converted data into a plurality of frequency packets of data;

determining a value of each frequency packet of data indicative of a feature of the packet of data; and

delivering at least one of the frequency packets and the values to a corresponding at least one neural network.

25. A method, as set forth in claim 16, wherein delivering the input data to a plurality of neural networks includes the steps of:

selecting desired channels of data from the input data;

analyzing interactions between the selected channels; and

delivering the analyzed channels of data to at least one neural network.

26. A method, as set forth in claim 25, wherein analyzing interactions between selected channels includes the step of applying a correlation algorithm to the selected channels.

27. A system for detecting and analyzing vibrations in a machine, comprising:

a test station operable to receive input data from a plurality of sensors, at least one sensor configured to detect characteristics of the machine indicative of machine vibration, and at least one other sensor configured to detect characteristics of the machine indicative of other than machine vibration;

a hub containing a plurality of neural networks, at least one neural network operable to receive and analyze vibration data, and at least one other neural network operable to receive and analyze non-vibration data, the hub further containing an expert system operable to receive analyzed data from the plurality of neural networks and responsively analyze vibration characteristics of the machine; and

a communications link operable to transmit data from the test station to the hub.