JPH0554280A - Method and device for alarm processing - Google Patents

Abstract

PURPOSE:To suppress alarms which are generated secondarily by outputting only important alarms among many alarms when abnormality occurs during plant operation and the alarms are outputted. CONSTITUTION:An important alarm decision means 6 decides whether or not causality information stored previously in a causality storage means 5 is generated as to an input signal by a causality decision means 4 and when suppression reason information indicated by the causality information to a generated alarm is generated, generated alarm is suppressed with the information and generated alarms which are not suppressed are outputted to a display means 7. In an on-line processing means 11, a suppression reason retrieval means 9 retrieves the suppression reason information regarding an alarm item specified by a causality information generating means 8 in a knowledge base 10 and outputs the result to the causality information generating means 8. The causality information which is thus generated is inputted from the causality information generating means 8 previously and stored in the causality means 5.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an alarm processing method and apparatus for plant operation monitoring, and more particularly to an alarm processing method and apparatus suitable for suppressing an alarm generated as a secondary event.

[0002]

2. Description of the Related Art In large-scale plants such as nuclear power plants, various systems are provided for safe and stable operation of the plants. The alarm processing system is
It is one of them, and it is an important system as a plant state display means when the plant is abnormal. For example, in a nuclear power plant, when a fault occurs in the plant, the alarm processing system may output hundreds of alarms depending on the degree of the fault. Even when a large number of alarms are generated in this way, the operator determines which alarm is the most important at that time.

In such a case, if it is possible to provide an alarm processing system that judges and displays a truly important alarm and does not display a secondary alarm (that is, suppresses a secondary alarm), The operator can easily understand the plant state, and the alarm processing system is very useful for the operator.

For example, in the prior art disclosed in Japanese Patent Laid-Open No. 64-12395, the knowledge about the attribute of the alarm, the cause of the alarm and the knowledge about the importance of the alarm are stored in the knowledge base, and the alarm is issued. Occasionally, the causal relationship of occurrence alarms is searched from this knowledge base to determine important alarms.

Further, according to the conventional technique described in Japanese Patent Laid-Open No. 63-282505, when a plant abnormality occurs, the cause event is determined, an alarm corresponding to the cause event is issued, and an operation corresponding to the cause event is performed. It is designed to guide the law to plant operators.

[0006]

DISCLOSURE OF THE INVENTION Problems to be Solved by the Invention
In the prior art described in Japanese Patent No. 395, regarding the occurrence alarm,
Searching for important alerts by reasoning mechanism. As described above, when the inference mechanism is operated to determine an important alarm, the processing time becomes long, and the important alarm is not output immediately after the alarm is issued, and several tens of seconds to several seconds have passed. After that, an important alarm will be output. This delay is not desirable because the operator needs to know the plant status immediately after the alarm is issued.
Further, in this conventional example, the importance of the alarm is predetermined, but since the importance may differ depending on the state of the plant, there is a problem that the importance cannot be predetermined for all alarms. ..

In the prior art disclosed in Japanese Patent Laid-Open No. 63-282505, the cause event is specified and the alarms other than the alarm corresponding to the cause event are suppressed as secondary alarms. But,
What is the causative event is usually found by shutting down the plant in which the abnormality occurred and then examining it,
It is difficult to identify the cause event when an abnormality occurs. Of course, if the cause event can be specified when an abnormality occurs and an alarm corresponding to it can be issued, it will be a very good alarm processing system, but at present, since the cause event is obtained by the inference mechanism as in the above conventional example, it is extremely Takes time to
There is a problem that it is not practical.

It is an object of the present invention to provide an alarm processing method and apparatus for quickly outputting important information and suppressing secondary alarms when an abnormality occurs in a plant.

[0009]

The above-mentioned object is to generate causal relationship information indicating a causal relationship with an alarm in advance based on alarm suppression reason information stored in a knowledge base and store it in a storage means to generate an alarm. This is sometimes achieved by determining an alarm that can be suppressed based on this causal relationship information and displaying the remaining alarms.

[0010]

Although it takes time to create the causal relationship information by the inference mechanism based on the knowledge base, in the present invention, the causal relationship information is created in advance by this inference mechanism and stored in the storage means. .. Then, when an alarm is issued, it is determined based on this causal relationship information whether or not the alarm should be suppressed, so that it is possible to quickly determine whether the alarm should be suppressed or output. Therefore, appropriate information can be promptly presented to the plant operator.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram of an alarm processing device according to an embodiment of the present invention. The alarm processing device of this embodiment, which monitors the state of a monitoring target 1 and issues an alarm when an abnormality occurs, is an alarm determination unit. 2, an event determination means 3, an important warning determination means 6, a display device 7, and an offline processing means 11 connected to the important warning determination means 6 offline. The important information judging means 6 comprises a causal relationship judging means 4 and a causal relationship storing means 5, and the offline processing means 11
It includes a causal relationship information creating means 8, a suppression reason searching means 9, and a knowledge base 10.

The monitoring object 1 outputs monitoring signals 1A, 1B, 1C consisting of process signals, equipment status signals, etc., and the important alarm judging means 6 is an alarm judging means based on the monitoring signal 1A from the monitoring object 1. Output signal of 2, monitoring target 1
The output signal of the event determination means 3 based on the monitor signal 1B from the monitor 1 and the monitor signal 1C from the monitor target 1 is fetched. The causal relationship determination means 4 of the important warning determination means 6 determines whether or not the causal relationship information stored in advance in the causal relationship storage means 5 has occurred, based on these input signals.
When it is determined that the suppression reason information indicated by the causal relationship information is generated with respect to the generated alarm, the generated alarm is suppressed with the information, and the generated alarm that is not suppressed is output to the display unit 7 as an important alarm. To do.

The causal relationship storage means 5 acquires and stores the causal relationship information created in advance by the causal relationship information creating means 8 of the off-line processing means 11. That is, the offline processing means 11 previously creates information that is a reason for suppression for each alarm item, and causes the alarm and the suppression reason to correspond one-to-one so that the causal relationship determination means 4 can easily determine the information. The causal relationship storage means 5 is used as the relationship information.
It is transferred to and stored in. For example, in the case of a nuclear power plant, there are about a few thousand alarm items, but since the systems that make up the plant are independent of each other, the alarms are also basically issued independently for each system. Therefore, the causal relationship of alarms is not established among all alarms, but basically, the causal relationship of alarms is established within the above system. It is possible to create and store causal relationship information indicating a relationship.

In the off-line processing means 11, first, the causal relationship information creating means 8 outputs the alarm item to the suppression reason searching means 9. The suppression reason search means 9 searches the knowledge base 10 for suppression reason information related to the designated alarm item, and outputs the result to the causal relationship information creation means 8.
The causal relationship information creating unit 8 organizes the suppression reason information for each alarm item and outputs it to the causal relationship storage unit 5 as causal relationship information indicating a causal relationship with the alarm. In this way, the causal relationship storage unit 5 can store the causal relationship information indicating the causal relationship with the alarm in advance. FIG. 7 is a configuration diagram of the causal relationship information stored in the causal relationship storage unit 5. As shown in FIG. 7, in the causal relationship information, the reason for suppressing the alarm item is defined in the form of a causal relationship table, and it is possible to determine whether or not the alarm generated should be suppressed by a lookup method. ing.

FIG. 2 is a flow chart showing an alarm processing procedure of the alarm processing apparatus shown in FIG. First, the important alarm judging means 6 takes in the input signals and judges whether or not there is an alarm generated among these signals. Causal relationship determination means 4
Searches the causal relationship table (FIG. 7) stored in the causal relationship storage means 5 to detect the information that is the reason for suppression, and determines whether or not the information has occurred. When the information that is the suppression reason is generated, it is determined whether or not the generation order of the generation alarm and the suppression reason information has a predetermined relationship. When the occurrence order has a predetermined relationship, the occurrence alarm is suppressed based on the suppression reason information. That is, since the alarm is a secondary alarm, no display output is made. When the order of occurrence does not have a predetermined relationship, it is determined that the occurrence alarm is important information and it is output as an alarm. If an alarm that is suppressed and is not output becomes a processing target again as a generated alarm in the next arithmetic processing, the number of alarm suppression processes increases as the number of generated alarms increases, which becomes an obstacle to high-speed processing. Therefore, a once-suppressed alarm is attached with a flag of "alarm generated in the past" so that it will not be treated as a generated alarm again until the suppression process is reset. As a result, if this flag is attached in the occurrence alarm detection processing, it is not considered as an occurrence alarm.

Further, if the suppression reason information for the alarm is not generated within a predetermined time, a message indicating that an abnormality has occurred in the corresponding device is created. For example, in the case where the interlock operation of the device works, an interlock is built so that the device B operates by the operation of the device A, and the alarm ANN-A indicating the operation of the device A and the device Alarm AN indicating the operation of B
It is assumed that NB is set. In this case, device A
Since the device B naturally operates as a result of the operation of, the alarm ANN-B should be inevitably generated after the alarm ANN-A is generated. Therefore, the alarm ANN
-When A occurs, an alarm AN is issued even if a certain time has elapsed after that.
If NB does not occur, it is abnormal. Therefore, in such a case, a message indicating that the alarm ANN-B is not issued is output and displayed. This message is output to the display device 7 together with the important warning.

The determination as to whether or not the information that is the reason for suppression in FIG. 7 has occurred is determined based on the following criteria. (1) With respect to the alarm signal, it is judged whether or not the specified alarm is generated. (2) With respect to the equipment state signal and the process signal, it is determined whether or not the state of each designated signal has a predetermined value. (3) The event is judged by whether or not the result obtained by the event judging means 3 is the specified event.

The flowchart of FIG. 2 will be described below in association with FIG. 7. First, it is assumed that the alarm name ANN-3 is detected by the detection process of the generated alarm. In the causal relationship table shown in FIG. 7, the alarm name ANN-1 is registered as the reason for suppressing the alarm name ANN-3. Next, with the alarm name ANN-3 being generated, this alarm name AN
When N-1 is generated, this alarm name ANN-3 is a secondary alarm generated by the occurrence of the alarm name ANN-1 which is the cause alarm. Since unnecessary information is given to the operator, this information (alarm name ANN-3) is suppressed and is not output. Therefore, cause alarm A, which is the reason for suppression
It is determined whether NN-1 has occurred. In this determination, if the alarm name ANN-1 has not occurred,
Next, it is monitored whether the alarm name ANN-1 is generated within a predetermined time. This monitoring is performed by the result generation monitoring counter. Here, if the alarm ANN- is given within a predetermined time,
If 1 occurs, the alarm ANN-1 does not need to output the alarm ANN-3. According to the causal relationship table shown in FIG. 7, the alarm ANN-1 further includes the alarm ANN-11, the device states ST-21-X, ST-22.
-X, process state PV-31-X, event EV-41 determines whether or not suppression is possible. Further, FIG. 7 shows that the alarm ANN-2 is suppressed by the generation of the alarm ANN-3.

Next, a method of creating the causal relationship table shown in FIG. 7 will be described. FIG. 3 is an explanatory diagram of a specific process of creating a causal relationship table. In FIG. 3, the knowledge of the alarm suppression reason information stored in the knowledge base 10 is represented by a frame structure as shown in FIG. 4, for example. The reason that the knowledge is expressed by the frame structure is that the description method is easy and the knowledge can be easily added or modified.

The causal relationship information creating means 8 activates the suppression reason searching means 9 to acquire the suppression reason information, and creates the causal relationship table shown in FIG. 7 based on this information. At this time, the search may be performed by designating an alarm name, but since it is easier to process if a predetermined alarm number is designated, the alarm number is designated here. The suppression reason searching means 9 searches the knowledge base 10 for alarm suppression reason information relating to the alarm specified by the alarm number, and returns the result to the causal relationship information creating means 8.

Next, the contents of these processes will be described. FIG. 5 shows a flowchart showing the respective processing procedures of the causal relationship information creating means 8 and the suppression reason searching means 9 in association with each other. The causal relationship information creating means 8 is shown in FIG.
As shown in (a), in step 1, the suppression reason search means 9
The alarm number to be searched is designated, and the suppression reason searching means 9 is activated in step 2. As a result, the suppression reason search means 9
Outputs the suppression reason information according to the processing shown in FIG. The causal relationship information creating means 8 fetches the suppression reason information output from the suppression reason searching means 9 in step 3. The above processing is carried out for all alarm items, and the causal relationship table shown in FIG. 7 is created.

Next, the processing of the suppression reason searching means 9 will be described using the knowledge example shown in FIG. In addition, this FIG.
In the alarm AN001, the alarm number is "1",
It indicates that the alarm that is the reason for suppression is AN005. The alarm AN005 indicates that the alarm number is "5". When the alarm number 1 specified by the causal relationship information creating unit 8 is fetched in step 1, the suppression reason searching unit 9 first searches all frame names and detects the frame name AN001 having the alarm number “1”. .. Next, in step 2, it is detected whether or not there is a cause alarm, a result alarm, an upper level alarm and a lower level alarm which are suppression reason information in this frame name AN001. In this example, the cause alarm AN005 is detected. Here, the upper alarm and the lower alarm are alarms that are set in multiple stages for one process amount, and for the changing direction of the process amount,
The upper level alarm is the upper level alarm. Cause alarm AN00
Since 5 indicates a frame name, in step 3, the classification number 5 of the cause alarm AN005 is detected and output as suppression reason information. In this example, there is only a cause alarm, but if there is another reason alarm, this is judged in step 4 and the same processing is performed. In addition, step 3
In FIG. 7, the state and event of the suppression reason information are also detected, which corresponds to the device state and process state and event shown in FIG.

As described above, all the suppression reason information for each alarm item is searched, and the suppression reason searching means 9 is searched.
To the causal relationship information creating means 8. In particular, in step 1, all frame names indicated by the alarm number are searched, so even if a plurality of alarm suppression reason information is distributively defined for one alarm item, these alarm suppression reason information are also defined. It is possible to retrieve all the suppression reason information corresponding to one alarm item. Therefore, even if there are multiple pieces of suppression reason information for one alarm item, FIG.
As shown in, it is possible to organize them by associating them with alarm items. Therefore, even if the knowledge about the suppression reason information for the alarm item is added, it is possible to retrieve all the alarm suppression reason information for the alarm item including the added alarm suppression reason information.

FIG. 6 shows an example of the configuration of an alarm processing device which is another embodiment of the present invention. 1 is different from the embodiment of FIG. 1 in that the causal relationship information created by the causal relationship information creating means 8 of the off-line processing means 11 is stored in the causal relationship storage means 5 of the important warning determination means 6 as means 81, 82 ,. This is the point where 51 is added. In FIG. 6, the causal relationship information created by the causal relationship information creating means 8 is converted into the causal relationship information 81.
It is stored in a portable information medium, for example, a floppy disk 81. This floppy disk 81
Take the causal relationship information stored in the floppy disk 81 to the causal relationship storing means 5 by taking the causal relationship information stored in the floppy disk 81. Output. With this configuration, even if the off-line means 11 is installed at a location physically different from the important alarm determination means 6, if the causal relationship information becomes new, this information is stored in the causal relationship storage means 5 as necessary. And can be stored easily.
The above portable information medium is a floppy
In addition to disks, tapes, ROM / ICs, IC cards, etc. can be considered.

In the above-mentioned alarm processing device, when an alarm is issued, the alarm that can be suppressed with respect to the generated alarm is determined by a lookup system based on the causal relationship information from the causal relationship storage means, and the remaining generated alarms are judged. Since the information is displayed, it is possible to suppress the secondary alarms and output only the important alarms immediately after the alarms are generated. Furthermore, as alarm suppression reason information, alarms that are suppression reasons, device statuses, process statuses, and events are assigned to each alarm item, so that alarms that occur secondarily among the alarms that occur according to the plant status are assigned. It can be efficiently suppressed. As a result, a truly important alarm can be provided immediately after the alarm is issued in the event of a plant abnormality, so that the plant operator can easily grasp the plant state, and the proper plant operation becomes possible.

[0026]

As described above, according to the present invention, when an abnormality occurs during plant operation, the output of a secondary alarm is suppressed, and only truly important information is quickly output to the plant operator. Since it can be presented, it is possible to support proper plant operation in the event of an abnormality.

[Brief description of drawings]

FIG. 1 is a configuration diagram of an alarm processing device according to a first embodiment of the present invention.

FIG. 2 is a flowchart showing an alarm processing procedure in the alarm processing device of FIG.

FIG. 3 is an explanatory diagram of creating a causal relationship table.

FIG. 4 is a diagram illustrating a knowledge representation of suppression reason information.

FIG. 5 is a flowchart showing a causal relationship table creation processing procedure.

FIG. 6 is a configuration diagram of an alarm processing device according to a second embodiment of the present invention.

FIG. 7 is a causal relationship table configuration diagram.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Monitored plant, 2 ... Alarm determination means, 3 ... Event determination means, 4 ... Causal relationship determination means, 5 ... Causal relationship storage means, 6 ... Important information determination means, 7 ... Display device, 8 ... Causal relationship information creation. Means, 9 ... Suppression reason search means, 10 ... Knowledge base, 11 ... Offline processing means.

Claims (16)

[Claims] 1. An alarm processing device for outputting an alarm when a monitored object becomes abnormal, in which causal relationship information indicating a causal relationship with an alarm is created in advance based on alarm suppression reason information stored in a knowledge base. A causal relationship information creating unit, a causal relationship storing unit that stores causal relationship information from the causal relationship information creating unit, and an alarm that can be suppressed based on the causal relationship information in the causal relationship storing unit when an alarm is issued. An alarm processing device comprising means for judging and displaying the remaining alarms. 2. An alarm processing device for outputting an alarm when a monitored object becomes abnormal, based on causal relationship information indicating a causal relationship with an alarm, which is stored in advance in a causal relationship storage means when the alarm occurs. An alarm processing device is provided with means for determining an alarm that can be suppressed among the generated alarms and displaying the remaining alarms. 3. A causal relationship storage means for inputting and storing causal relationship information indicating a causal relationship with an alarm in an alarm processing device that outputs an alarm when a monitored object becomes abnormal.
An alarm processing device comprising means for determining an alarm that can be suppressed based on causal relationship information from the causal relationship storage means when an alarm is issued and displaying the remaining alarms. 4. An alarm processing device that outputs an alarm when an object to be monitored becomes abnormal, and whether or not suppression reason information indicated by causal relationship information indicating a causal relationship with the alarm is generated with respect to the generated alarm. An alarm processing device comprising: a judgment means for judging, and a means for suppressing an occurrence alarm and displaying the remaining occurrence alarms based on the result of the judgment means. 5. The means for suppressing an occurrence alarm according to claim 4, when the occurrence order of the occurrence alarm and the causal relationship information indicating the causal relationship between the occurrence alarm has a predetermined relationship. An alarm processing device characterized by being provided. 6. The means according to claim 4, wherein the suppression reason information indicated by the causal relationship information indicating the causal relationship with the alarm is generated within a predetermined time after the specific alarm is generated. And a means for displaying information indicating that the suppression reason information does not occur within a predetermined time when the suppression reason information does not occur within a predetermined time. 7. The causal relationship information indicating a causal relationship with an alarm according to any one of claims 1 to 6, which is alarm reason information for each alarm item, as a suppression reason information, a device status, a process status,
An alarm processing device characterized by assigning events. 8. The suppression reason for each alarm item according to any one of claims 1 to 7, wherein a means for designating each alarm item for all alarm items and activating a suppression reason information is activated from the knowledge base for each alarm item. An alarm processing device comprising means for extracting information. 9. The alarm suppression reason information to be stored in the knowledge base according to claim 1, which is described in advance in a frame representation, searches a frame indicating the alarm for each alarm item, and suppresses the alarm defined in the frame. An alarm processing device, characterized in that it comprises means for detecting a frame of reason information and extracting suppression reason information shown in the frame for each alarm item. 10. An alarm processing device that outputs an alarm when a monitored object becomes abnormal, converts alarm suppression reason information stored in a knowledge base into causal relationship information indicating a causal relationship with the alarm, and converts the information. Based on the means for storing the information in a portable information medium, the causal relationship storage means for extracting the causal relationship information from the information medium and storing the information, and the causal relationship information from the causal relationship storage means when an alarm is issued. An alarm processing device comprising means for determining an alarm that can be suppressed and displaying the remaining alarms. 11. An alarm processing device for outputting an alarm when a monitored object becomes abnormal, wherein alarm suppression reason information stored in a knowledge base is described in advance in a frame representation and a frame indicating the alarm for each alarm item. To create a causal relationship information indicating a causal relationship with the alarm by detecting the suppression reason information frame defined in the frame, extracting the suppression reason information shown in the frame for each alarm item, Means, a causal relationship storage means for storing the causal relationship information from the creating means, and a warning alarm that can be suppressed based on the causal relationship information from the causal relationship storage means when an alarm is issued, and the remaining occurrence alarms are issued. An alarm processing device provided with a display means. 12. In an alarm processing method for outputting an alarm when a monitored object becomes abnormal, causal relationship information indicating a causal relationship with an alarm is created in advance based on alarm suppression reason information stored in a knowledge base. Then, when an alarm is issued, an alarm processing method is characterized in that an alarm that can be suppressed is determined based on this causal relationship information and the remaining alarms are displayed. 13. An alarm processing method for outputting an alarm when a monitored object becomes abnormal, based on causal relationship information indicating a causal relationship with an alarm stored in advance in a causal relationship storage means when the alarm is issued. The alarm processing method is characterized by determining which of the generated alarms can be suppressed and displaying the remaining generated alarms. 14. An alarm processing method for outputting an alarm when a monitored object becomes abnormal, wherein causal relationship information indicating a causal relationship with the alarm is stored in a storage device, and the causal relationship information is stored when the alarm is issued. An alarm processing method characterized in that an alarm that can be suppressed is determined first and the remaining alarms are displayed. 15. In an alarm processing method for outputting an alarm when a monitored object becomes abnormal, whether or not suppression reason information indicated by causal relationship information indicating a causal relationship with the alarm is generated with respect to the generated alarm. An alarm processing method, which comprises making a judgment, suppressing the generated alarms based on the result of the judgment, and displaying the remaining generated alarms. 16. The alarm processing method according to claim 12, wherein the causal relationship information is created based on an interlock of a monitoring target.