Dynamic environmental change compensation of sensor data in structural health monitoring systems

• US 8,036,836 B2
• Filed: 04/16/2008
• Issued: 10/11/2011
• Est. Priority Date: 04/16/2007
• Status: Active Grant

• Current Assignee: Acellent Technologies, Inc.
• Original Assignee: Acellent Technologies, Inc.
• Inventors: Chang, Fu-Kuo, Beard, Shawn J., Liu, Bao
• Primary Examiner(s): Dunn; Drew A
• Assistant Examiner(s): CHERRY, STEPHEN J
• Application Number:
  US12/104,354
• Publication Number:
  US 20080255774A1
• Time in Patent Office: 1,273 Days
• Field of Search: 702 33- 36
• US Class Current: 702/33
• CPC Class Codes: G01N 2291/0427 Flexural waves, plate waves, e.g. Lamb waves, tuning fork, cantilever
• CPC Class Codes: G01N 2291/106 one or more transducer arrays
• CPC Class Codes: G01N 29/041 on the surface of the material, e.g. using Lamb, Rayleigh or shear waves
• CPC Class Codes: G01N 29/36 Detecting the response signal , e.g. electronic circuits specially adapted therefor
• CPC Class Codes: G01N 29/4463 Signal correction, e.g. distance amplitude correction [DAC], distance gain size [DGS], noise filtering

Abstract

A method for adjusting signal data detected in a structural health monitoring (SHM) system to compensate for the effects of environmental variables acting thereon includes constructing a baseline data space comprised of sets of signal data. Current signal data sets are collected for comparison to the baseline data space. The collected current signal data sets are amended to best match baseline signal data sets in the baseline data space. A set of indices are computed for comparing the amended current signal data set to the baseline signal data sets. A threshold for detection is determined by outlier detection for the computed indices. A signal in the collected signal data set is determined to be detected on the basis of the threshold. A representation of the detected signal strength is provided on the basis of the computed indices.

Claims

1. A method for adjusting sensor signal data detected in a structural health monitoring (SHM) system to compensate for the effects of an environmental variable acting on the data, the method comprising:

detecting one or more sets of a sensor signal data from an array of transducers mounted on a structure being monitored by the SHM system;

constructing a baseline data space comprising the one or more sets of the sensor signal data detected;

detecting one or more current sets of the sensor signal data from the transducers at a later time at which the environmental variable is acting to affect the data;

amending the baseline data space with the current sensor signal data sets while they are being detected; and

,

determining if the current sensor signal exceeds a selected threshold value on the basis of the amended baseline data space; and

repeating the detecting one or more current sets of the sensor signal data, the amending, and the determining for each of a plurality of signal propagation paths in the structure respectively extending between associated pairs of the transducers.

2. The method of claim 1, wherein the constructing of the baseline data space comprises:

selecting a propagation path;

collecting a selected number of sensor signal data sets corresponding to the selected path; and

,

repeating the preceding steps for each of the propagation paths of the structure.

3. The method of claim 2, wherein the constructing of the baseline data space further comprises removing electromagnetic interference (EMI) cross-talk from each of the sensor signal data sets detected.

4. The method of claim 3, wherein the constructing of the baseline data space further comprises:

computing a Euclidean distance parameter between all pairs of the signal data sets for the selected path; and

,

selecting a distance parameter from among the computed distance parameters as a threshold distance parameter d_fTH.

5. The method of claim 4, wherein the constructing of the baseline data space further comprises sorting the sensor signal data, wherein the sorting of the signal data comprises:

clustering the signal data sets in a selected number of groups according to a mutual proximity of the signal data sets as determined by the computed distance parameter; and

,

defining a center for each clustered group of signal data sets as the location of a signal data set within the cluster group having the smallest average value of distance parameter to all other members of that cluster group.

6. The method of claim 5, further comprising:

discarding all signal data sets in each respective cluster group except the signal data set at the defined group center; and

,

including in each of the baseline signal data sets only the respective group center baseline signal data set thereof.

7. The method of claim 6, further comprising forming the baseline data space from the plurality of baseline data sets.

8. The method of claim 7, further comprising determining the value of the signal detection threshold for each propagation path.

9. The method of claim 8 wherein the determining of the threshold value comprises:

selecting a first baseline signal data set from among the group center baseline signal data sets;

calculating a clustered baseline data index I_BL(j) for all baseline signal data sets relative to the first baseline signal data set for the selected path from the relationship;

10. The method of claim 4, wherein the computing of the Euclidean distance parameter pairs of signal data sets for all pairs of baseline signal data sets for the selected path comprises:

calculating the distance parameter d^i,j_f from the relationship;

11. The method of claim 1, wherein the amending of the baseline data space comprises:

detecting a current sensor signal data set for a selected propagation path of the structure;

holding the current signal data set in a baseline buffer;

matching the current data set to a best baseline signal data set;

specifying a threshold value for the detection of a signal based on the best match baseline data set; and

,

determining whether the current data set represents a detection of a local event comprising either an environmental effect or a damage of the structure on the basis of the threshold value specified.

12. The method of claim 11, wherein the matching of the current signal data set to a best baseline signal data set comprises:

removing electromagnetic interference (EMI) cross-talk from the current signal data set;

retrieving the plurality of baseline signal data sets for the selected path;

segmenting the current signal data set and the plurality of baseline signal data sets to select a first arrival signal;

computing a cross correlation between the current data set and each of the plurality of baseline signal data sets;

obtaining a time shift for each of the cross correlation computations corresponding to each of the plurality of baseline signal data sets;

selecting the baseline signal data set that has the minimum difference to the respective shifted current signal as the best match baseline; and

,

shifting the current signal data set waveform by the time shift corresponding the to best match baseline data set.

13. The method of claim 11, further comprising:

detecting the local spatial event using outlier detection techniques;

specifying a survival time T_s corresponding to the detected local event;

measuring the persistence of the local event;

determining that the local event is damage if the persistence time is greater than T_s; and

,

determining that the local event is an environmental effect if the persistence time is less than T_s.

14. The method of claim 13, wherein the detecting of the local spatial event using outlier detection techniques comprises:

computing an index of the shifted current data set relative to the best match baseline date for each of the plurality of selected paths;

computing a mean value and a standard deviation from the plurality of best match damage indices;

computing a Grubbs test value z(k) for each best match damage index from the relationship;

15. The method of claim 13, further comprising;

adding the current data of the local event-related paths to the baseline data space if the persistence time is less than T_s;

deleting the current data from the baseline buffer if the persistence time is less than T_s;

deleting the current data from the baseline buffer if the persistence time is greater than T_s;

determining whether the damage indices are larger than a selected threshold;

determining whether actual damage has been detected using non-destructive testing (NDT) techniques or by visual inspection; and

,

updating the baseline data set with the current data if actual damage has not been detected.