Analyzing signals generated by rotating machines using an order mask to select desired order components of the signals

• US 20020046006A1
• Filed: 10/09/2001
• Published: 04/18/2002
• Est. Priority Date: 04/19/2000
• Status: Active Grant

• Current Assignee: National Instruments Corporation (Emerson Electric Company)
• Original Assignee: National Instruments Corporation (Emerson Electric Company)
• Inventors: Jin, Wei, Shao, Hui, Qian, Shie
• Granted Patent: US 6,477,472 B2
• Time in Patent Office: Days
• US Class Current: 702/183
• CPC Class Codes: G01H 1/003 of rotating machines G01H1/10 takes precedence

Abstract

A signal analysis system and method for analyzing an input signal acquired from a mechanical system. The mechanical system may include at least one rotating apparatus. The signal analysis system may be configured to: (a) receive samples of the input signal, (b) perform an invertible joint time-frequency transform (e.g., a Gabor transform) on the samples of the input signal to produce a first array of coefficients which depend on time and frequency, (c) construct an order mask array in response to user input designed to select coefficients from the first array which correspond to one or more order components in the input signal, (d) mask the first array of coefficients with the order mask array to generate a second array of coefficients which depend on time and frequency, wherein the second array of coefficients correspond to the one or more order components in the input signal, (e) generate a time domain signal from the second array of coefficients, and (f) present the time domain signal to a user on a presentation device. The signal analysis system may generate the time domain signal from the first coefficients by performing an inverse joint time-frequency transform on the first coefficients. The time domain signal may then be useable in analyzing the mechanical system.

Claims

1. A method for analyzing one or more order components from an input signal acquired from a mechanical system containing a rotating apparatus, the method comprising:

receiving samples of the input signal;

performing an invertible joint time-frequency transform on the samples of the input signal to produce a first array of coefficients which depend on time and frequency;

constructing an order mask array in response to user input;

masking the first array of coefficients with the order mask array to generate a second array of coefficients which depend on time and frequency;

generating a time domain signal from the second array of coefficients; and

presenting the time domain signal to a user on a presentation device;

wherein the time domain signal is useable in analyzing the mechanical system.

2. The method of claim 1, wherein the input signal is sampled in time.

3. The method of claim 1, wherein the input signal comprises a plurality of order components, and wherein each of the order components comprises a rotation-periodic signal present in the input signal.

4. The method of claim 3, wherein said masking the first array of coefficients with the order mask array operates to select one or more desired order components from the input signal;

wherein the second array of coefficients correspond to the one or more desired order components.

5. The method of claim 4, wherein the order mask array comprises first values at locations corresponding to the one or more desired order components and second values at other locations.

6. The method of claim 4, wherein the order mask array comprises one values at locations corresponding to the one or more desired order components and zero values at other locations.

7. The method of claim 4, wherein the order mask array comprises an array of coefficients similar in structure to the first array;

wherein the order mask array comprises one values at locations corresponding to the one or more desired order components and zero values at other locations.

8. The method of claim 4, the method further comprising:

displaying a visual representation of the first array of coefficients, wherein the visual representation indicates the plurality of order components; and

generating the order mask array based on user input to the visual representation.

9. The method of claim 8, further comprising:

receiving user input which indicates selection of one or more points in the visual representation, wherein said user input indicates the one or more desired order components; and

generating the order mask array based on the user input to the visual representation.

10. The method of claim 9, further comprising:

determining one or more time-frequency curves corresponding to the one or more points, wherein the order mask array is generated based on the one or more time-frequency curves.

11. The method of claim 1, wherein said presentation device comprises at least one speaker;

wherein said presenting the time domain signal to the user on the presentation device comprises audially presenting the time domain signal as sound output from the at least one speaker.

12. The method of claim 1, wherein said presentation device comprises a display;

wherein said presenting the time domain signal to the user on the presentation device comprises visually displaying the time domain signal on the display.

13. The method of claim 1, wherein said generating the time domain signal from the modified array of coefficients comprises performing an inverse joint time-frequency transform on the modified array of coefficients.

14. The method of claim 1, wherein the joint time-frequency transform comprises the Gabor transform.

15. The method of claim 1, further comprising adjusting the mechanical system in response to analysis of the time domain signal.

16. The method of claim 15, wherein said adjusting the mechanical system comprises one or more of:

1) changing a design of the mechanical system;

     2) replacing one or more components of the mechanical system;

     3) predicting a failure of one or more components of the mechanical system; and

     4) adding a sound-absorbent material to one or more locations of the mechanical system.

17. The memory medium of claim 1, wherein the input signal comprises a signal generated by a microphone proximate to the mechanical system.

18. The method of claim 1, wherein said constructing an order mask comprises:

computing an instantaneous rotation frequency signal with respect to time;

multiplying the instantaneous rotation frequency signal by one or more order numbers to generate one or more order curves;

generating a third array of coefficients, wherein coefficients of the third array equal a first value in one or more time-frequency neighborhoods which contain the one or more order curves respectively, wherein the order mask comprises the third array, wherein the coefficients of the third array equal a second value outside the one or more time-frequency neighborhoods.

19. The method of claim 18, wherein said computing the instantaneous rotation frequency signal comprises:

receiving a first signal comprising a stream of arrival times, wherein each pair of successive arrival times in said stream indicates a constant angular displacement of said rotating apparatus;

filtering the arrival times to determine a stream of derivative estimates, wherein said stream of derivative estimates are used to determine the instantaneous rotation frequency signal.

20. The method of claim 19, wherein said filtering the arrival times to determine a stream of derivative estimates comprises filtering the arrival times with an FIR filter.

21. The method of claim 20, wherein coefficients of said FIR filter are determined by operating on a linear system derived from a Taylor series expansion of time with respect to rotation angle of the rotating element;

wherein matrix coefficients of the linear system have the form

$\frac{{\left(n<br><br>\text{\hspace{1em}<br><br>}<br><br>\mathrm{\Delta <br><br>\theta <br><br>}\right)}^k}{k!},$

 

where index k ranges from 1 to an upper limit M, wherein index k ranges through a set of M distinct integer values different from zero.

22. The method of claim 19, wherein said computing the instantaneous rotation frequency signal further comprises reciprocating the stream of derivative estimates.

23. The method of claim 18, wherein the one or more time-frequency neighborhoods have frequency bandwidths controlled by a second user input.

24. The method of claim 18, computing the one or more order numbers in response to user input.

25. The method of claim 24, wherein said computing the one or more order numbers in response to the user input comprises:

displaying a visual representation of the first array of coefficients on a display screen;

receiving location data in response to one or more user selection events specifying one or more locations on the display screen;

computing one or more time-frequency pairs in response to the location data;

dividing the frequency of each time-frequency pair by a value of the instantaneous rotation frequency signal corresponding to the time of the time-frequency pair.

26. The method of claim 24, wherein said computing the one or more order numbers in response to the user input comprises:

resealing the first array of coefficients to generate a fourth array of coefficients, wherein the resealed fourth array of coefficients depend on order and time;

displaying a visual representation of the fourth array of coefficients on a display screen;

receiving location data in response to one or more user selection events specifying one or more locations on the display screen.

28. The memory medium of claim 27, wherein the input signal is sampled in time.

29. The memory medium of claim 27, wherein the input signal comprises a plurality of order components, and wherein each of the order components comprises a rotation-periodic signal present in the input signal.

30. The memory medium of claim 29, wherein said masking the first array of coefficients with the order mask array operates to select one or more desired order components from the input signal;

wherein the second array of coefficients correspond to the one or more desired order components.

31. The memory medium of claim 30, wherein the order mask array comprises first values at locations corresponding to the one or more desired order components and second values at other locations.

32. The memory medium of claim 30, wherein the order mask array comprises one values at locations corresponding to the one or more desired order components and zero values at other locations.

33. The memory medium of claim 30, wherein the order mask array comprises an array of coefficients similar in structure to the first array;

wherein the order mask array comprises one values at locations corresponding to the one or more desired order components and zero values at other locations.

34. The memory medium of claim 30, wherein the program instructions are further executable to implement:

displaying a visual representation of the first array of coefficients, wherein the visual representation indicates the plurality of order components; and

generating the order mask array based on user input to the visual representation.

35. The memory medium of claim 34, wherein the program instructions are further executable to implement:

receiving user input which indicates selection of one or more points in the visual representation, wherein said user input indicates the one or more desired order components; and

generating the order mask array based on the user input to the visual representation.

36. The memory medium of claim 35, wherein the program instructions are further executable to implement:

determining one or more time-frequency curves corresponding to the one or more points, wherein the order mask array is generated based on the one or more time-frequency curves.

37. The memory medium of claim 27, wherein said presentation device comprises at least one speaker;

wherein said presenting the time domain signal to the user on the presentation device comprises audially presenting the time domain signal as sound output from the at least one speaker.

38. The memory medium of claim 27, wherein said presentation device comprises a display;

wherein said presenting the time domain signal to the user on the presentation device comprises visually displaying the time domain signal on the display.

39. The memory medium of claim 27, wherein said generating the time domain signal from the modified array of coefficients comprises performing an inverse joint time-frequency transform on the modified array of coefficients.

40. The memory medium of claim 27, wherein the joint time-frequency transform comprises the Gabor transform.

41. The memory medium of claim 27, wherein the mechanical system is operable to be adjusted in response to analysis of the time domain signal.

42. The memory medium of claim 27, wherein the input signal comprises a signal generated by a microphone proximate to the mechanical system.

43. The memory medium of claim 27, wherein said constructing an order mask comprises:

computing an instantaneous rotation frequency signal with respect to time;

multiplying the instantaneous rotation frequency signal by one or more order numbers to generate one or more order curves;

generating a third array of coefficients, wherein coefficients of the third array equal a first value in one or more time-frequency neighborhoods which contain the one or more order curves respectively, wherein the order mask comprises the third array, wherein the coefficients of the third array equal a second value outside the one or more time-frequency neighborhoods.

27. A memory medium comprising program instructions for analyzing one or more order components from an input signal acquired from a mechanical system containing a rotating apparatus, wherein the program instructions are executable to implement:

receiving samples of the input signal;

performing an invertible joint time-frequency transform on the samples of the input signal to produce a first array of coefficients which depend on time and frequency;

constructing an order mask array in response to user input;

masking the first array of coefficients with the order mask array to generate a second array of coefficients which depend on time and frequency;

generating a time domain signal from the second array of coefficients;

presenting the time domain signal to a user on a presentation device;

wherein the time domain signal is useable in analyzing the mechanical system.

44. A memory medium comprising at least one software program for analyzing one or more order components from an input signal generated by a mechanical system containing a rotating apparatus, wherein the memory medium stores:

samples of the input signal;

the first array of coefficients which depend on time and frequency, wherein the first array of coefficients is generated by performing an invertible joint time-frequency transform on the samples of the input signal;

an order mask array which is operable to be applied to the first array of coefficients to generate a second array of coefficients which depend on time and frequency;

the second array of coefficients; and

a time domain signal generated from the second array of coefficients;

wherein the time domain signal is operable to be presented to a user on a presentation device;

wherein the time domain signal is useable in analyzing the mechanical system.

45. The memory medium of claim 44, wherein the memory medium further stores:

at least one software program operable to;

perform the invertible joint time-frequency transform on the samples of the input signal to produce the first array of coefficients which depend on time and frequency;

generate the order mask array;

apply the order mask array to the first array of coefficients to generate the second array of coefficients; and

generate the time domain signal from the second array of coefficients.