Method and system for monitoring the condition of an individual machine

• US 20010005821A1
• Filed: 12/27/2000
• Published: 06/28/2001
• Est. Priority Date: 12/23/1999
• Status: Active Grant

• Current Assignee: ABB Schweiz AG (ABB Limited)
• Original Assignee: ABB AB (ABB Limited)
• Inventors: Ottosson, Torbjorn
• Granted Patent: US 6,694,286 B2
• Time in Patent Office: Days
• US Class Current: 702/185
• CPC Class Codes: G05B 23/0235 based on a comparison with predetermined threshold or range, e.g. "classical methods", carried out during normal operation; threshold adaptation or choice; when or how to compare with the threshold
• CPC Class Codes: G05B 23/0237 based on parallel systems, e.g. comparing signals produced at the same time by same type systems and detect faulty ones by noticing differences among their responses
• CPC Class Codes: G07C 3/00 Registering or indicating the condition or the working of machines or other apparatus, other than vehicles

Abstract

A method for monitoring the condition of an individual machine belonging to a selected class of machines, wherein operational data significant for that condition are compared with a reference data according to a predetermined criterion, and wherein an output signal indicating the result of the comparison is formed, the method comprising the steps of:

creating a first set of operational data for the selected class of machines on the basis of data stored in a reference database,

creating a second set of operational data for the individual machine based on measurements performed on that machine,

forming a first reference data in dependence on values of at least a selection of said first set of operational data,

forming a second reference data in dependence on values of said second set of operational data,

comparing said second reference data with said first reference data in accordance with the first predetermined criterion, and

forming the output signal in dependence on the result of said comparison.

Claims

1. A method for monitoring the condition of an individual machine belonging to a selected class of machines, wherein operational data significant for that condition is compared with a reference data according to a predetermined criterion, and wherein an output signal indicating the result of the comparison is formed, characterised in that the method comprises the steps of:

creating a first set of operational data for the selected class of machines on the basis of data stored in a reference database,

creating a second set of operational data for the individual machine based on measurements performed on that machine,

forming a first reference data in dependence on values of at least a selection of said first set of operational data,

forming a second reference data in dependence on values of said second set of operational data,

comparing said second reference data with said first reference data in accordance with the first predetermined criterion, and

forming the output signal in dependence on the result of said comparison.

2. A method according to

claim 1

, characterised in that it comprises the steps of:

uploading said second set of operational data to said reference database (10, 30), and

storing it in said reference database (10, 30).

3. A method according to

claim 2

, characterised in that it comprises the steps of performing a non-fault check on said second set of operational data before said step of uploading and storing said second set of operational data in said reference database.

4. A method according to any of the preceding claims, characterised in that the step of creating said second set of operational data comprises the step of storing said second set of operational data in an individual database, located at the reference database.

5. A method according to any of the preceding claims, characterised in that said step of forming a first reference data comprises the step of selecting in said reference database operational data for a selected group of parameters and forming said first reference data in dependence on operational data for said selected group of parameters, and that said step of forming a second reference data comprises the step of selecting in said second set of operational data operational data for said selected group of parameters.

6. A method according to the preceding claims, characterised in that it comprises the steps of

performing at least one statistical analysis according to a predetermined statistical method of the operational data stored in said reference database (10, 30), and

creating the first reference data in dependence on the result of said statistical analysis.

7. A method according to any of the preceding claims, characterised in that the reference database (10, 30) contains a time series of values for one or more parameters of the operational data for every individual machine that can be connected to the reference database (10, 30).

8. A method according to any of the preceding claims, characterised in that the first reference data as provided by the reference data base (10, 30) remains unaffected by the measured operational data of the individual machine, at least until actions on the machine itself, e.g. service measures or component exchanges, have been carried out.

9. A method according to any of the preceding claims characterised in that the second reference data is formed at the local machine.

10. A method according to any of the preceding claims characterised in that the second reference data is formed centrally in a system including the reference database (10, 30).

11. A method according to any of the preceding claims, characterised in that the first reference data is compared with the second reference data at the local machine.

12. A method according to

claim 11

, characterised in that the first reference data is formed from a copy of the individual database, located at the local machine.

13. A method according to any of the preceding claims, characterised in that the first reference data is compared with the second reference data centrally in a system including the reference database (10, 30).

14. A method according to patent

claim 13

, characterised in that the second reference data is obtained from the individual database, located at the system, including the reference database (10, 30).

15. A method according to any of the preceding claims, characterised in that two or more first reference data are each formed in dependence on the operational data collected from the selected class of machines.

16. A method according to any of the preceding claims, characterised in that two or more second reference data are each formed in dependence on the operational data collected from the individual machine.

17. A method according patent

claim 6

, characterised in that the statistical method used is a multivariate analysis.

18. A method according patent

claim 17

, characterised in that the statistical method used includes multivariate analysis projection techniques.

19. A method according patent

claim 18

, characterised in that the multivariate analysis is a Principal Component Analysis (PCA).

20. A method according patent

claim 18

, characterised in that the multivariate analysis is a Projection to Latent Structures (PLS).

21. A method according to patent

claim 3

, characterised in that the step of performing a non fault check on said second set of operational data before said steps of uploading and storing, is followed by discarding operational values in excess of predetermined value limits from the operational data subsequently updated to the central database.

22. A method according to

claim 21

, characterised in that the step of discarding operational values in excess of predetermined value limits is followed by collecting discarded values in an additional database for error modelling and/or failure mode modelling.

23. A method according to patent

claim 2

, characterised in that downloading reference data to the individual machine and/or uploading to the reference database (10, 30) is carried out at predetermined points of time.

24. A method according to any of the preceding claims, characterised in that the operational data measured for the individual machine is compared with operational data previously collected for the same machine.

25. A method according to any of the preceding claims, characterised in that changes over time in the operational data of the individual machine are compared with changes over time in the reference data formed from a plurality of similar machines.

26. A method according to any of the preceding claims, characterised in that the age of the data contained in the reference database (10, 30) is evaluated, and the data having reached a prescribed age is compressed accordingly by reducing the depth of information detail.

27. System for the determination of the condition of an individual machine, with a transducer (16) on each individual object;

means for data acquisition (18) connected with the transducer (16); and

a pre-processing unit (20) connected to the acquisition means (18), for comparing the transducer data with a reference data;

characterised in that a reference database (30) containing measured operational data for a selected class of machines is provided, to which the individual machine can be connected for transferring transducer (16) data and receiving at least one first reference data.

28. System according to patent

claim 27

, characterised in that the reference database (10, 30) provides storage (28) for the transducer data of each individual machine that can be connected to the reference database (10, 30).

29. System according to patent

claim 28

, characterised in that the reference database (10, 30) comprises an individual database in which the transducer data of each individual machine is stored.

30. System according to patent

claim 29

, characterised in that a copy of the individual database in which the transducer data of each individual machine is stored is arranged locally at the individual machine.

31. System according to any of patent claims 27-30, characterised in that the reference database (10, 30) is provided with an analysis tool for multivariate modelling, for creating statistical models based upon the data gathered in the reference database (30) and the at least one first reference data.

32. System according to patent

claim 27

-31, characterised in that the individual machine is an electrical machine.

33. System according to any of patent claims 27-32, characterised in that the pre-processing unit (20) is capable of executing the pre-processing in at least two modes, mode 1 being a model comparison with transducer (16) data for a supervision of the object, and mode 2 being a raw data filtering, intended for data modelling or focusing.

34. The use of a system according to any of the patent claims 27-33 for monitoring the condition of an individual machine.

35. The use of a system according to any of the patent claims 27-33 for monitoring the condition of a selected class of machines.