Integrated acoustic leak detection processing system

• US 5,533,383 A
• Filed: 08/18/1994
• Issued: 07/09/1996
• Est. Priority Date: 08/18/1994
• Status: Expired due to Fees

• Current Assignee: General Electric Company
• Original Assignee: General Electric Company
• Inventors: Greene, David A., Greene, Rosemary A., Gaubatz, Donald C.
• Primary Examiner(s): Williams, Hezron E.
• Assistant Examiner(s): LARKIN, DANIEL SEAN
• Application Number:
  US08/292,656
• Time in Patent Office: 691 Days
• Field of Search: 73/40.5 A, 73/587, 73/592, 376/250, 376/251, 376/252, 395/22, 364/507, 364/550
• US Class Current: 73/40.50A
• CPC Class Codes: G01M 3/24 using infrasonic, sonic, or ultrasonic vibrations
• CPC Class Codes: G01N 2291/02836 Flow rate, liquid level
• CPC Class Codes: G01N 2291/106 one or more transducer arrays
• CPC Class Codes: G01N 29/036 by measuring frequency or resonance of acoustic waves
• CPC Class Codes: G01N 29/14 using acoustic emission techniques echo of particles G01N29/046; measuring mechanical vibrations or acoustic waves in solids in general G01H1/00

Abstract

A system for mapping absolute acoustic noise intensity in a three-dimensional acoustic noise field, and using three-dimensional absolute noise intensities to infer operational or performance characteristics of components or structures within the monitored field. Localized noise sources are extracted using a distant array of transducers, and the absolute intensity can be measured even when totally masked by background noise at the transducer locations. The system includes an integrated sensor installation; a neural network detection system algorithm; a zoom system to precisely examine a small region in the steam generator vessel; a fuzzy logic system detection algorithm; and an expert detection system. The signal processing subsystem operates at three different levels of detection. At the top level of detection a trained neural network system monitors the vessel length for any indication of a leak. The proposed approach uses discriminators which do not require any beamforming of the sensor signals. The presence of a leak and the general location in the vessel is determined using a fuzzy logic expert system. The leak indications are passed to a second level of detection: a beamformed system which will monitor the indicated portion of the vessel. A third level of detection systematically monitors the tagged locations to determine if a leak is really present, remaining on the specific locations to detect a leak above the defined threshold to a high degree of accuracy. The actions to be taken will be controlled by a second fuzzy logic expert interface controller.

Claims

1. An apparatus for detecting an anomaly in a monitored volume, comprising:

a plurality of transducers coupled to a respective plurality of localized portions of said monitored volume, each of said transducers outputting a respective analog electrical signal having a low-frequency component which is a function of the vibrational displacement of said localized portions of said monitored volume;

means for converting said respective analog electrical signals into respective digital electrical signals;

means for multiplexing said respective digital electrical signals from said respective transducers; and

neural network processing means for recognizing a predetermined pattern in the digital electrical signals of any transducer having said predetermined pattern masked by background noise, said predetermined pattern reflecting a possible anomaly in said monitored volume.

2. The apparatus as defined in claim 1, wherein said analog electrical signals output by said transducers have high-frequency components which are a function of the ultrasonic waves propagating through said respective one of said plurality of localized portions of said monitored volume, further comprising a plurality of means for splitting the respective outputs of said plurality of transducers into said low- and high-frequency components.

3. The apparatus as defined in claim 2, wherein said neural network processing means comprises a first neural network processor for recognizing predetermined patterns in the digital electrical signals corresponding to said low-frequency components of said analog electrical signals and a second neural network processor for recognizing predetermined patterns in the digital electrical signals corresponding to said high-frequency components of said analog electrical signals.

4. The apparatus as defined in claim 3, further comprising chemical sensor means for monitoring for the presence in said monitored volume of a chemical compound produced when an anomaly of a predetermined kind is present in said monitored volume, and fuzzy logic anomaly detection processing means coupled to receive the outputs of said first and second neural network processors and said chemical sensor means, said fuzzy logic anomaly detection processing means being programmed in accordance with a first set of fuzzy logic rules to output predetermined commands in response to a combination of inputs thereto which indicate an anomaly, said inputs including at least said outputs of said first and second neural network processors and said chemical sensor means.

5. The apparatus as defined in claim 4, further comprising beamforming means for monitoring a selected set of transducers in the probable location of said indicated anomaly in response to said predetermined commands from said fuzzy logic anomaly detection processing means.

6. The apparatus as defined in claim 5, further comprising fuzzy logic interface processing means for controlling the operating conditions within said monitored volume by applying a second set of fuzzy logic rules to data output by said beamforming means as the result of said monitoring of said selected set of transducers and data output by said fuzzy logic anomaly detection processing means.

7. The apparatus as defined in claim 6, wherein said monitored volume is the outer shell of a sodium-heated steam generator, and said fuzzy logic interface processing means reduces the power being produced by said steam generator in response to inputs thereto indicating an anomaly inside said monitored volume.

8. The apparatus as defined in claim 3, wherein said monitored volume is a vessel having a wall, and each of said transducers comprises an acoustic transducer, further comprising a plurality of rigid attachment means, each of said rigid attachment means coupling a respective one of said plurality of acoustic transducers to the external surface of a respective one of a plurality of localized portions of said vessel wall lying along a helix.

9. The apparatus as defined in claim 3, further comprising fuzzy logic anomaly detection processing means coupled to receive the outputs of said first and second neural network processors, said fuzzy logic defect detection processing means being programmed in accordance with a first set of fuzzy logic rules to output predetermined commands in response to a combination of inputs thereto which indicate an anomaly, said inputs including at least said outputs of said first and second neural network processors.

10. The apparatus as defined in claim 9, further comprising beamforming means for monitoring a selected set of transducers in the probable location of said indicated anomaly in response to said predetermined commands from said fuzzy logic anomaly detection processing means.

11. The apparatus as defined in claim 9, wherein said analog-to-digital converting means have a changeable sampling rate.

12. The apparatus as defined in claim 10, further comprising fuzzy logic interface processing means for controlling operating conditions within said monitored volume by applying a second set of fuzzy logic rules to data output by said beamforming means as the result of said monitoring of said selected set of transducers and data output by said fuzzy logic anomaly detection processing means.

13. A method for preventing damage caused by an anomaly in a monitored volume, comprising the steps of:

detecting acoustic waves produced by said anomaly at a plurality of locations in said monitored volume;

detecting the amount of a chemical compound at a location inside said monitored volume, said chemical compound being produced when an anomaly of a predetermined kind is present in said monitored volume;

applying a set of fuzzy logic rules to at least the data acquired by said detecting steps; and

controlling the operating conditions within said monitored volume in dependence on the results of said step of applying fuzzy logic rules to at least said acquired data.

14. The method as defined in claim 13, wherein said monitored volume is the outer shell of a sodium-heated steam generator, the power produced by said steam generator being reduced in response to results of said step of applying fuzzy logic rules to at least said acquired data indicating an anomaly inside said monitored volume.

15. An apparatus for detecting an anomaly in a monitored volume, comprising:

a plurality of transducers coupled to a respective plurality of localized portions of said monitored volume, each of said transducers outputting a respective analog electrical signal having a low-frequency component which is a function of the vibrational displacement of said localized portions of said monitored volume;

means for converting said respective analog electrical signals into respective digital electrical signals;

beamforming means for combining said digital electrical signals corresponding to acoustic waves originating from a predetermined unit volume inside said monitored volume at the same instant in time; and

fuzzy logic processing means for controlling the operating conditions inside said vessel by applying a set of fuzzy logic rules to at least the data output by said beamforming means.

16. The apparatus as defined in claim 15, wherein said analog-to-digital converting means have a changeable sampling rate.

17. The apparatus as defined in claim 15, wherein said monitored volume is the outer shell of a sodium-heated steam generator, and said fuzzy logic processing means reduces the power being produced by said steam generator in response to inputs thereto indicating an anomaly inside said monitored volume.

18. A method for locating an anomaly within a monitored volume, comprising the step of:

coupling a plurality of transducers to a plurality of localized portions of said monitored volume so that each of said plurality of transducers outputs an analog electrical signal having a low-frequency component which is a function of the vibrational displacement of a respective one of said plurality of localized portions of said monitored volume and a high-frequency component which is a function of the vibrational/ultrasonic waves propagating through said respective one of said plurality of localized portions of said monitored volume;

converting said respective analog electrical signal components into respective digital electrical signals;

processing said digital electrical signals through a neural network trained to recognize predetermined patterns in said respective digital electrical signals, said predetermined patterns reflecting possible anomalies within said monitored volume;

monitoring for the presence in said monitored volume of a chemical compound produced when an anomaly of a predetermined kind is present in said monitored volume; and

processing the results of said neural network processing and said monitoring in accordance with a first set of fuzzy logic rules to output predetermined commands in response to results indicating an anomaly.

19. The method as defined in claim 18, further comprising the step of beamforming the low-frequency components of a subset of said plurality of transducers nearest to the probable location of said indicated anomaly in response to said predetermined commands.

20. The method as defined in claim 19, further comprising the step of controlling the operating conditions inside said monitored volume by applying a second set of fuzzy logic rules to at least the result of said beamforming and the results of said processing in accordance with said first set of fuzzy logic rules.