System and method for measuring the operation of a device

• US 5,406,502 A
• Filed: 06/29/1993
• Issued: 04/11/1995
• Est. Priority Date: 06/29/1993
• Status: Expired due to Fees

• Current Assignee: Elbit Ltd (IDB Holding Corporation Limited)
• Original Assignee: Elbit Ltd (IDB Holding Corporation Limited)
• Inventors: Haramaty, Yoav, Lewin, Daniel R.
• Primary Examiner(s): Voeltz, Emanuel T.
• Assistant Examiner(s): Wachsman, Hal D.
• Application Number:
  US08/084,687
• Time in Patent Office: 651 Days
• Field of Search: 364/551.01, 364/550, 364/184, 364/580, 395/906, 395/907, 395/914, 395/915, 395/575, 324/248, 356/436
• US Class Current: 702/183
• CPC Class Codes: G05B 19/4063 Monitoring general control system G05B19/4062 takes precedence
• CPC Class Codes: Y10S 706/906 Process plant

Abstract

A system and method for measuring the operation of a device is disclosed. A plurality of virtual sensors which define a multi-dimensional virtual sensor space sense the operation of the device. A normal operation determiner determines at least one region of normal operation for the device. An abnormal condition determiner provides a warning when the operation does not fall within the normal operation region which is defined within the multi-dimensional virtual sensor space. A maintenance method and system are also disclosed for maintaining at least one device. The maintenance system includes a measurement system and an abnormality processor. The measurement system measures the operation of the device and provides an indication of when the device is not operating normally. The abnormality processor predicts, based on the output of the measurement system, when the device is likely to fail.

Claims

1. A measurement system for measuring the operation of a device, the measurement system comprising:

a plurality of virtual sensors wherein ranges of output from said plurality of virtual sensors define a multi-dimensional virtual sensor space;

a normal operation determiner, coupled to said plurality of virtual sensors at least during installation, for defining at least one multi-dimensional normal operation region from at least one datapoint within said multi-dimensional virtual sensor space, said at least one datapoint being defined from output of said plurality of virtual sensors received at least during said installation; and

an abnormal condition determiner, coupled to said plurality of virtual sensors during a period of regular operation, for providing an abnormality indication when output of said plurality of virtual sensors during said period of regular operation does not fall within said at least one multi-dimensional normal operation region.

2. A measurement system according to claim 1 and wherein said at least one multi-dimensional normal operation region is at least one quantized region around said at least one datapoint within said multi-dimensional virtual sensor space.

3. A measurement system according to claim 1 and additionally comprising an operation space definer for defining a reduced dimension operation space from said multi-dimensional virtual sensor space, wherein said reduced dimension operation space has fewer axes than said multi-dimensional virtual sensor space.

4. A measurement system according to claim 3 and additionally comprising a region definer for defining, from datapoints within said multi-dimensional virtual sensor space said at least one multi-dimensional normal operation region as a collection of distances of said datapoints from said axes of said reduced dimension operation space.

5. A measurement system according to claim 1 and wherein said plurality of virtual sensors comprise physical sensors and computed sensors, wherein said computed sensors are sensors whose output is a function of output of at least one physical sensor.

6. A maintenance system for maintaining at least one device, the system comprising:

a region and path storage unit for storing;

(i) at least one multi-dimensional normal operation region defined by at least one multi-dimensional state previously defined by an operator as indicating normal operation of said at least one device; and

(ii) a list of failure paths, each failure path ending in at least one different failure and each failure path having a multiplicity of multi-dimensional nodes, wherein each multi-dimensional node is defined by a multi-dimensional abnormal operation region defined by at least one multi-dimensional state previously defined by said operator as indicating abnormal operation of said at least one device;

a measurement system comprising;

measurement means for measuring the operation of said at least one device;

comparison means for comparing output of said measurement means with said at least one multi-dimensional normal operation region and for providing an abnormality indication when said output of said measurement means does not fall within at least one of said at least one multi-dimensional normal operation region; and

an abnormality processor for determining when said abnormality indication falls within at least one of said multi-dimensional nodes of said failure paths and for predicting therefrom when said at least one device is likely to fail.

7. A maintenance system according to claim 6 and wherein said measurement means comprises a plurality of virtual sensors wherein ranges of output from said plurality of virtual sensors define a multi-dimensional virtual sensor space;

and further wherein said measurement system also comprises;

a normal operation determiner, coupled to said plurality of virtual sensors at least during installation, for defining said at least one multi-dimensional normal operation region from at least one datapoint within said multi-dimensional virtual sensor space, said at least one datapoint being defined from output of said plurality of virtual sensors received at least during said installation; and

an abnormal condition determiner, coupled to said plurality of virtual sensors during a period of regular operation, for providing an abnormality indication when output of said plurality of virtual sensors during said period of regular operation does not fall within said at least one multi-dimensional normal operation region.

8. A maintenance system according to claim 7 and wherein said abnormality processor additionally comprises a predictor for predicting when said at least one device is likely to fail based on which multi-dimensional node on which of said failure paths said abnormality indication fell and on how long said at least one device took to reach said multi-dimensional node.

9. A maintenance system according to claim 7 and wherein said abnormality processor further comprises a region rebuilder operative to rebuild said at least one multi-dimensional normal operation region when said abnormality indication is falsely produced.

10. A measurement system according to claim 7 and wherein said at least one multi-dimensional normal operation region is at least one quantized region around said at least one datapoint within said multi-dimensional virtual sensor space.

11. A maintenance system according to claim 7 and additionally comprising an operation space definer for defining a reduced dimension operation space from said multi-dimensional virtual sensor space, wherein said reduced dimension operation space has fewer axes than said multi-dimensional virtual sensor space.

12. A measurement system according to claim 11 and additionally comprising a region definer for defining, from datapoints within said multi-dimensional virtual sensor space said at least one multi-dimensional normal operation region as a collection of distances of said datapoints from said axes of said reduced dimension operation space.

13. A method of measuring the operation of a device, the method comprising the steps of:

defining a multi-dimensional virtual sensor space from ranges of output data of a plurality of virtual sensors;

determining at least one multi-dimensional normal operation region from at least one datapoint within said multi-dimensional virtual sensor space, said at least one datapoint being defined from output data of said plurality of virtual sensors received at least during an installation period; and

providing an abnormality indication when output data of said plurality of virtual sensors during a period of regular operation does not fall within said at least one multi-dimensional normal operation region.

14. A method according to claim 13 and wherein said at least one multi-dimensional normal operation region is at least one quantized region around said at least one datapoint within said multi-dimensional virtual sensor space.

15. A method according to claim 13 and wherein said step of determining comprises the step of defining a reduced dimension operation space from said multi-dimensional virtual sensor space, wherein said reduced dimension operation space has fewer axes than said multi-dimensional virtual sensor space.

16. A method according to claim 15 and wherein said step of determining additionally comprises the step of defining, from datapoints within said multi-dimensional virtual sensor space said at least one multi-dimensional normal operation region as a collection of distances of said datapoints from said axes of said reduced dimensional operation space.

17. A method according to claim 13 and wherein said plurality of virtual sensors comprise physical sensors and computed sensors, wherein said computed sensors are sensors whose output is a function of output of at least one physical sensor.

18. A method for maintaining at least one device comprising the steps of:

storing;

(i) at least one multi-dimensional normal operation region defined by at least one multi-dimensional state previously defined by an operator as indicating normal operation of said at least one device; and

(ii) a list of failure paths, each failure path ending in at least one different failure and each failure path having a multiplicity of multi-dimensional nodes, wherein each multi-dimensional node is defined by a multi-dimensional abnormal operation region defined by at least one multi-dimensional state previously defined by said operator as indicating abnormal operation of said at least one device;

measuring the operation of said at least one device;

comparing output of said step of measuring with said at least one multi-dimensional normal operation region;

providing an abnormality indication when said output of said step of measuring does not fall within at least one of said at least one multi-dimensional normal operation region; and

verifying when said abnormality indication falls within at least one of said multi-dimensional nodes of said failure paths and predicting therefrom when said at least one device is likely to fail.

19. A method according to claim 18 and wherein said step of measuring comprises the steps of:

defining a multi-dimensional virtual sensor space from ranges of output data of a plurality of virtual sensors;

determining at least one multi-dimensional normal operation region from at least one datapoint within said multi-dimensional virtual sensor space, said at least one datapoint being defined from output data of said plurality of virtual sensors received at least during an installation period; and

providing an abnormality indication when output of said plurality of virtual sensors during a period of regular operation does not fall within said at least one multi-dimensional normal operation region.

20. A method according to claim 19 and wherein said step of verifying and predicting additionally comprises the step of producing a prediction of when said at least one device will fail based on which multi-dimensional node on which of said failure paths said abnormality indication fell and on how long said at least one device took to reach said multi-dimensional node.

21. A method according to claim 19 and wherein said step of verifying and predicting further comprises the step of rebuilding said at least one multi-dimensional normal operation region when said abnormality indication is falsely produced.

22. A method according to claim 19 and wherein said at least one multi-dimensional normal operation region is at least one quantized region around said at least one datapoint within said multi-dimensional virtual sensor space.

23. A method according to claim 19 and wherein said step of determining comprises the step of defining a reduced dimension operation space from said multi-dimensional virtual sensor space, wherein said reduced dimension operation space has fewer axes than said multi-dimensional virtual sensor space.

24. A method according to claim 23 and wherein said step of determining additionally comprises the step of defining, from datapoints within said multi-dimensional virtual sensor space said at least one multi-dimensional normal operation region as a collection of distances of said datapoints from said axes of said reduced dimension operation space.