Method for pre-detecting abnormality sign of nuclear power plant device including processor for determining device importance and warning validity, and system therefor

Patent metadata

Field	Value
Publication number	US-11200790-B2
Application number	US-201716328627-A
Country	US
Kind code	B2
Filing date	Aug 29, 2017
Priority date	Aug 29, 2016
Publication date	Dec 14, 2021
Grant date	Dec 14, 2021

Abstract

The present invention provides a method for early warning of an abnormality sign of a device, which includes device importance and warning validity determination, and a system therefor. The method for early warning of the abnormality sign of the device comprises: a first step of determining by an early warning processing apparatus whether a device monitoring signal value exceeds a normal operation range by using a weight value on the basis of monitoring-parameter-specific importance data which has been previously analyzed by an operator; a second step of generating a warning by the early warning processing apparatus when the device monitoring signal value exceeds the normal operation range; and a third step of determining by a warning determination apparatus whether the generated warning is a valid warning, which is subject to a warning analysis and to be traced.

First claim

The invention claimed is:

1. A method for early warning of an abnormality sign of a device, which includes device importance and warning validity determination, the method comprising:
a first step of determining, by an early warning processing device, whether a device monitoring signal value exceeds a normal operation range by using a weight value based on importance data for each monitoring variable which has been previously analyzed by an operator;
a second step of generating a warning, by the early warning processing device, when the device monitoring signal value exceeds the normal operation range; and
a third step of determining, by the warning determination device, whether the generated warning is a valid warning, which is subject to a warning analysis and to be traced.

2. The method for early warning of the abnormality sign of the device of claim 1 , wherein the first step includes:
performing analogue-digital conversion, by a data processing device, of the monitoring signals received from measurement instruments; and calculating a current value of the monitoring signals and transmitting the current value to the early warning processing device and an importance categorizing device in a predetermined time interval.

3. The method for early warning of the abnormality sign of the device of claim 2 , further comprising searching, by the importance categorizing device, an importance grade and the weight value of the monitoring variable to which the transmitted current value corresponds in an importance database, and transmitting the importance grade and the weight value to the early warning processing device.

4. The method for early warning of the abnormality sign of the device of claim 3 , further comprising calculating, by the early warning processing device, a residual by using the transmitted current value, the weight value and a preconfigured pattern learning model.

5. The method for early warning of the abnormality sign of the device of claim 4 , further comprising determining, by the early warning processing device, whether the residual passes the normal operation range which is preconfigured by an operator.

6. The method for early warning of the abnormality sign of the device of claim 1 , wherein the third step includes receiving, by the warning determination device, current state information of the warning and categorizing the current state information into an announced warning and the warning to be analyzed.

7. The method for early warning of the abnormality sign of the device of claim 6 , wherein the announced warning corresponds to at least one of a warning generated by power plant main control room, a warning generated by a device replacement operation, and a warning generated by an experiment or a maintenance scheduled in advance.

8. The method for early warning of the abnormality sign of the device of claim 6 , if the warning is the warning be analyzed, when it is determined that a power plant maintenance is required as a result of the warning analysis and trace by the operator, wherein the warning determination device notices the generation of warning to at least one of a power plant operator computer and a power plant operator personal wireless communication device.

9. A system for early warning of an abnormality sign of a device, which includes device importance and warning validity determination, the system comprising:
an early warning processing device for determining whether a device monitoring signal value exceeds a normal operation range by using a weight value based on importance data for each monitoring variable which has been previously analyzed by an operator, and generating a warning when the device monitoring signal value exceeds the normal operation range; and
a warning determination device for determining whether the generated warning is a valid warning, which is subject to a warning analysis and to be traced.

10. The system for early warning of the abnormality sign of the device of claim 9 , further comprising a data processing device for performing analogue-digital conversion of the monitoring signals received from measurement instruments, and calculating a current value of the monitoring signals and transmitting the current value to the early warning processing device and an importance categorizing device in a predetermined time interval.

11. The system for early warning of the abnormality sign of the device of claim 10 , further comprising the importance categorizing device for searching an importance grade and a weight value of the monitoring variable to which the transmitted current value corresponds in an importance database, and transmitting the importance grade and the weight value to the early warning processing device.

12. The system for early warning of the abnormality sign of the device of claim 11 , wherein the early warning processing device calculates a residual by using the transmitted current value, the weight value and a preconfigured pattern learning model.

13. The system for early warning of the abnormality sign of the device of claim 9 , wherein the warning determination device receives current state information of the warning and categorizes the current state information into an announced warning and an warning be analyzed.

14. A method for early detection of an abnormality sign of a device of a nuclear power plant, the method comprising:
performing, by a pattern learning device, a pattern learning by using past device monitoring data of a predetermined period and grouping monitoring variables showing a similar pattern for a same device; and
generating, by a prediction value calculation device, a prediction value of a current monitoring data received in real time by using patterns of the grouped monitoring variables.

15. The method for early detection of the abnormality sign of the device of claim 14 , wherein the step of grouping monitoring variables includes:
a first step of grouping the monitoring variables for each of same devices;
and
a second step of grouping the grouped monitoring variables for each of the same devices showing the similar pattern.

16. The method for early detection of the abnormality sign of the device of claim 15 , wherein the step of generating a prediction value calculates the prediction value by comparing the current monitoring data with a pattern learning value of the similar monitoring variable group belonged to the same device and providing a higher weight value as the more the current monitoring data are similar to the pattern learning value.

17. The method for early detection of the abnormality sign of the device of claim 14 , wherein the past device monitoring data and the current monitoring data are digital data going through analogue-digital conversion.

18. The method for early detection of the abnormality sign of the device of claim 14 , wherein the past device monitoring data and the current monitoring data are data related to early warning subject monitoring variable which is predetermined by an operator.

19. The method for early detection of the abnormality sign of the device of claim 14 , further comprising generating, by an early warning processing device, an early warning by obtaining a residual using an absolute value of a difference between the prediction value and the current data, when the residual is greater than a normal operation range residual.

20. The method for early detection of the abnormality sign of the device of claim 19 , wherein the normal operation range residual is stored in database, which is designated

Assignees

• Korea Hydro & Nuclear Power Co

Inventors

• Min Ji Ho
• Kim Ji In
• Ha Hooh Kwon
• Oh Jeong Hag
• Park Young Sheop
• Park Chi Yong
• Kim Dae Woong
• Kim Yang Seok
• Lee Byoung Oh

Classifications

• G08B21/187

  Machine fault alarms · CPC title

• G21D3/04

  Safety arrangements (emergency protection of reactor G21C9/00) · CPC title

• G21D3/001 Primary

  Computer implemented control · CPC title

• G06N20/00

  Machine learning · CPC title

• G21C17/00

  Monitoring; Testing (measuring in general G01); {Maintaining} · CPC title