MACHINE CONDITION INDICATION SYSTEM

• US 20080106424A1
• Filed: 01/11/2008
• Published: 05/08/2008
• Est. Priority Date: 01/31/2005
• Status: Active Grant

• Current Assignee: Computational Systems Incorporated (Emerson Electric Company)
• Original Assignee: CSI Technology Incorporated (Emerson Electric Company)
• Inventors: Bouse, Kai, Bowers, Stewart, Robinson, Jim, Reeves, Todd, Piety, Richard
• Granted Patent: US 8,174,402 B2
• Time in Patent Office: Days
• US Class Current: 340/635
• CPC Class Codes: G05B 19/4065 Monitoring tool breakage, life or condition
• CPC Class Codes: G05B 2219/31467 Display of operating conditions of machines, workcells, selected programs
• CPC Class Codes: G05B 2219/32297 Adaptive scheduling, feedback of actual proces progress to adapt schedule
• CPC Class Codes: G05B 23/0232 based on qualitative trend analysis, e.g. system evolution

Abstract

A machine condition monitor resides in close proximity to a machine train, such as an AC inductive electric motor coupled to a driven unit such as a centrifugal pump or fan. A plurality of sensing devices are connected to the machine and the machine condition monitor. Machine state and condition parameters sensed by these devices are utilized by the machine condition monitor to derive machine condition values for each component and for the machine train as a whole. These derived condition values are transmitted via an industrial communications network to a control center where they can be trended and monitored. With no particular knowledge of machine analysis techniques or severity of specific machine faults, and no special training, Operators and Production Planners can use the condition values to plan production schedules, adjust process parameters, and request appropriate maintenance action.

Claims

1. A method for determining an operational condition of a machine, comprising:

(a) accepting data representing sensor signals acquired from a plurality of sensors attached to the machine, where the sensor signals are indicative of operational characteristics of the machine;

(b) processing the data to calculate a plurality of characteristic parameters indicative of the operational condition of the machine; and

(c) operating on the characteristic parameters to derive a single numerical condition value which is indicative of the operational condition of the machine.

2. The method of claim 1 wherein step (b) further comprises processing the data to calculate a plurality of wave-form-based characteristic parameters.

3. The method of claim 1 wherein step (b) further comprises processing time waveform data to calculate frequency spectral data using discrete Fourier transform techniques.

4. The method of claim 3 further comprising processing the frequency spectral data to calculate a plurality of frequency-based characteristic parameters.

5. The method of claim 4 further comprising processing the frequency spectral data to calculate the plurality of frequency-based characteristic parameters using order-based techniques dependent on the rotational speed of the machine.

6. The method of claim 1 wherein step (b) further comprises processing the plurality of characteristic parameters using arithmetic capabilities and domain-specific intrinsics based on configurable calculation descriptions to express arbitrarily complex relationships between the characteristic parameters and to calculate derived characteristic parameters to be used in determining the condition of the machine.

7. The method of claim 1 wherein step (b) further comprises maintaining cascaded hierarchical statistical histories of the characteristic parameters.

8. The method of claim 7 further comprising querying the cascaded hierarchical statistical histories of the characteristic parameters to ascertain historical similarity based on minimum, maximum, mean, median, moments, and Z-score relative to specified temporal groupings.

9. The method of claim 7 further comprising querying the cascaded hierarchical statistical histories of the characteristic parameters to characterize parameter populations according to correlation, regression, trend fitting and projection, and statistical means analysis.

10. The method of claim 1 wherein step (b) further comprises processing the data using analysis techniques selected from the group consisting of windowing, averaging, interpolation and decimation, static and dynamic filters, convolution, peak group signatures, energy distributions, extreme low-frequency, digital integration and differentiation, cross-channel wobble and eccentricity, cross-channel transfer functions, joint time-frequency, cepstrum, wavelets, demodulation, and PeakVue®

analytical techniques.

11. The method of claim 1 wherein step (c) further comprises operating on the characteristic parameters to derive the single numerical condition value as an analog value between 0.0 and 1.0, where 0.0 indicates a failed condition and 1.0 indicates a proper operating condition.

12. A method for reporting an operational condition of a machine to a central monitoring system using a machine condition monitoring device, the method comprising:

(a) attaching a plurality of sensors at a plurality of locations on the machine for sensing operational characteristics at the plurality of locations;

(b) producing a plurality of sensor signals using the sensors, the sensor signals indicative of the operational characteristics;

(c) acquiring data from the plurality of sensor signals related to the operational characteristics;

(d) processing the data to calculate a plurality of characteristic parameters indicative of the operational condition of the machine;

(e) operating on the characteristic parameters to derive a single numerical condition value which is indicative of the operational condition of the machine; and

(f) communicating the numerical condition value to the central monitoring system.

13. The method of claim 12 wherein step (e) further comprises operating on the characteristic parameters to derive the single numerical condition value as an analog value between 0.0 and 1.0, where 0.0 indicates a failed condition and 1.0 indicates a proper operating condition.

14. A machine condition indication device configured for attachment to a machine for reporting information regarding the operational condition of the machine to a central monitoring system, the device comprising:

a plurality of sensors mounted at a plurality of locations on the machine for sensing operational characteristics at the plurality of locations and for producing a plurality of sensor signals indicative of the operational characteristics;

a main processor coupled to the plurality of sensors for receiving the plurality of sensor signals, for acquiring data from the plurality of sensor signals related to the operational characteristics, processing the data to calculate a plurality of characteristic parameters indicative of the operational condition of the machine, and operating on the characteristic parameters to derive a single numerical condition value which is indicative of the operational condition of the machine.

15. The machine condition indication device of claim 14 further comprising a communication interface for communicating the numerical condition value via a communication link to the central monitoring system.

16. The machine condition indication device of claim 15 wherein the communication interface is further for communicating the numerical condition value via a fieldbus network.

17. The machine condition indication device of claim 14 further comprising the main processor for acquiring time waveform data from one or more of the sensors and processing the time waveform data to calculate the plurality of characteristic parameters.

18. The machine condition indication device of claim 17 further comprising the main processor for acquiring the waveform data from more than one of the sensors simultaneously.

19. The machine condition indication device of claim 14 further comprising:

a digital signal processor coupled to a memory subsystem and analog-to-digital converters for receiving the data from one or more of the plurality of sensor signals; and

an interface for transferring the data to the main processor.

20. The machine condition indication device of claim 19 wherein the interface further comprises a switched bank of dynamic random access memory.

21. The machine condition indication device of claim 14 further comprising the main processor for processing the data to calculate a plurality of waveform-based characteristic parameters selected from the group consisting of mean, minimum, maximum, span, absolute peak, variance, skewness, kurtosis, standard deviation, crest factor, autocorrelation, and time synchronous correction.

22. The machine condition indication device of claim 14 further comprising the main processor for processing time waveform data to calculate frequency spectral data using discrete Fourier transform techniques selected from the group consisting of single-frequency DFT and fast Fourier transform.

23. The machine condition indication device of claim 22 further comprising the main processor for processing the frequency spectral data to calculate a plurality of frequency-based characteristic parameters selected from the group consisting of peak, phase, power bands, energy bands, synchronous energy families, non-synchronous energy, and sideband energy families.

24. The machine condition indication device of claim 23 further comprising the main processor for processing the frequency spectral data to calculate the plurality of frequency-based characteristic parameters using order-based techniques dependent on the rotational speed of the machine.

25. The machine condition indication device of claim 14 further comprising the main processor for processing the plurality of characteristic parameters using arithmetic capabilities and domain-specific intrinsics based on configurable calculation descriptions to express arbitrarily complex relationships between the characteristics, where the configurable calculations descriptions are downloadable to the device over the communications interface.

26. The machine condition indication device of claim 14 further comprising:

the main processor for taking N number of data acquisitions from the plurality of sensor signals, for processing the data from the N number of data acquisitions to calculate N baseline values for each of the characteristic parameters, for calculating a representative baseline pair for each of the characteristic parameters based upon the N number of values, where each baseline pair <

A_base, σ

_base>

comprises a baseline value, A_base, and a baseline sigma, σ

_base,

the main processor for taking N number of new data acquisitions from the plurality of vibration sensor signals subsequent to taking the initial baseline data acquisitions, for processing the data from the N number of new data acquisitions to calculate N number of new values for each of the characteristic parameters, for calculating a new pair for each of the characteristic parameters based upon the N number of new values, where each new pair <

A_new, σ

_new>

comprises a new value, A_new, and a new sigma, σ

_new, and for operating on the baseline and new values and sigmas according to;

if (A_new>

Min_diag) and (A_new<

A_base) then A_base=A_new
and
if (σ

_new>

Min_diag) and (σ

_new<

σ

_base) then σ

_base=σ

_new

where Min_diag is a minimum amplitude value of the characteristic parameter which must be exceeded for machine condition analysis to be performed and which varies according to the operating environment of the machine.

27. The machine condition indication device of claim 14 further comprising the main processor for taking data acquisitions from the plurality of sensor signals, for processing the data from the data acquisitions to calculate instantaneous values, x_i, for each of the characteristic parameters, and for determining whether each instantaneous value, x_i, of the characteristic parameter for the data acquisition satisfies the following:

$\begin{array}{cc}{x}_i\notin <br><br>\left[\frac{A_{\mathrm{base}}}{N_{\left[E\right]}},A_{\mathrm{base}}·<br><br>N_{\left[E\right]}\right]& \left(1\right)\end{array}$
x_i∉

└

A_base−

M_[E]σ

_base,A_base+M_[E]σ

_base┘

 

 

(2)

where N_[E] is a parametric amplitude scaling factor and M_[E] is a parametric sigma scaling factor which define a width of a stable region around the baseline value according to the operating environment of the machine.

28. The machine condition indication device of claim 14 further comprising a data storage device for storing configuration values, characteristic parameter calculation descriptions, fault similarity inference tables, current baseline data, and histories of the characteristic parameters.