JPH0712982A - Processing apparatus for alarm at plant - Google Patents

Abstract

PURPOSE:To provide a plant alarm processing apparatus which indicates an operator a cause of alarm, a proceeding course of an abnormal phenomenon, a predicting alarm and countermeasures when a plant alarm is continuously generated. CONSTITUTION:The apparatus is provided with a knowledge database 6, a monitoring means 2, a diagnosing means 3, an output processing means 4 and a controlling means 5. The knowledge database 6 stores the attribute, cause/result, countermeasures, degree of importance and generating conditions for each alarm. Moreover, the causal relationship between alarms is stored in the database 6. The monitoring means 2 forms an alarm state data including the state and time when the alarm is generated. The diagnosing means 3 determines a cause alarm of an abnormality, a predicting alarm and a countermeasures guide. The monitoring result by the monitoring means 2 and the diagnosing result by the diagnosing means 3 are output from the output processing means 4. The start/stop of the operation of the apparatus is controlled by the controlling means 5. The monitoring means 2 is activated by the controlling means 5 to monitor a plant every predetermined cycle, and the diagnosing means 3 is actuated by the monitoring means 2 to make diagnosis. The output processing means 4 is started to display by the monitoring means 2 and diagnosing means 3.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to the processing of alarms in large-scale plants such as nuclear power plants, thermal power plants, chemical plants, etc., especially when a large number of alarms occur continuously in a short time. The present invention relates to a plant alarm processing device that identifies the cause and result of an event and displays the countermeasure to the operator.

[0002]

2. Description of the Related Art Generally, in a large-scale plant such as a nuclear power plant, a thermal power plant or a chemical plant, when any abnormality occurs, the abnormality is prevented from expanding or various safety maintenance is performed to positively converge the abnormal event. Means are provided. As one of these means for maintaining safety, there is a plant interlock mechanism that combines the operations of various relays according to the control logic.

This plant interlock mechanism controls the operation of component equipment of the plant in the direction of canceling the abnormal event when the abnormal event occurs. Normally, along with the operation of the interlock mechanism, an alarm indicating the operation of the device due to this is displayed on the control panel of the plant. In addition to the alarm by the interlock mechanism, an alarm indicating physical behavior of the plant such as operation of components of the plant and pressure / flow rate is displayed on the plant control panel.

When an abnormal event occurs, the operator can quickly grasp the state of the plant by the above various alarms,
Appropriate measures need to be taken. However, if the plant is large or the anomalous event is in transition,
Many alarms are continuously displayed, and it is often impossible to grasp the plant status depending on the partial alarm display. In this case, the operator has to determine the cause of the abnormality from all the lit alarms, or to determine the direction of the abnormal event progress according to the sequence of the lit alarms and take appropriate action.

It is a heavy burden for the operator to make such a judgment and work when an urgent abnormality occurs, and requires abundant experience and knowledge. In order to reduce the burden on the operator and prevent misjudgment, it has been required to develop a plant alarm processing device that supports the operator's situation judgment in the event of an abnormality.

On the other hand, when an alarm is generated during the rated output operation of the plant, an alarm processing device is proposed which judges the cause alarm of the abnormality and the important alarm and displays a predictive alarm that will occur in the near future to the operator. Has been done.

According to this alarm processing device, the operator can understand the cause of the abnormal event from the cause alarm and can take appropriate action in consideration of the displayed countermeasures. The time burden is greatly reduced, and the possibility of misjudgment is greatly reduced.

[0008]

However, the above-mentioned conventional plant alarm processing apparatus is in the rated output operation,
Moreover, since it is configured so that it does not operate except when an abnormality occurs, there is little opportunity for the operator to come into contact with the operation of this alarm processing device, and the fact that the operator is not accustomed to the device reduces the reliability of the device. It was a cause.

Further, although the above-mentioned conventional plant alarm processing device can process an alarm during the rated output operation of the plant,
It was not possible to process the alarms that occur during the plant operation mode such as the plant safety stop after the abnormal convergence of the plant, the restart after the safety stop, and the normal start.

In this conventional device, the alarms other than during the rated output operation are left to the operator's judgment as in the prior art, so the burden on the operator during these operations cannot be reduced. Considering that an abnormality is likely to occur in an operation mode other than the rated output operation and an abnormal event progresses quickly, the conventional plant alarm processing device described above does not sufficiently reduce the burden on the operator.

Therefore, an object of the present invention is to solve the problems of the conventional plant alarm processing apparatus described above, to monitor the plant alarms and equipment states in all operation modes of the plant, and when alarms occur continuously, It is to provide a plant alarm processing device that specifies a cause alarm that causes a cause, a path of an abnormal event, and a predicted alarm that will occur in the near future, and also displays corresponding measures to an operator.

[0012]

In order to achieve the above object, a plant alarm processing apparatus according to the present invention includes an alarm attribute including an alarm name and a determination condition, an alarm generation cause, and an alarm generation result. For each alarm, knowledge about the countermeasures against the alarm, the importance of the alarm evaluated by the influence on the plant, and the alarm generation condition for the cause of the alarm is stored, and the knowledge of causal relation between the alarms is developed in a network. A stored knowledge database, monitoring means for taking in plant alarms and control signals, detecting an occurrence of an abnormality by the knowledge database, and creating alarm state data including an alarm generation state and an occurrence time; and the alarm state data. Diagnostic means that identifies the path of abnormality development from the above, determines the cause alarm of the abnormality, the predicted alarm that will occur next, and the corresponding action guide, and the operation A non-operation alarm when a device to be operated does not operate, an alarm progress path when an alarm occurs, a cause alarm, a prediction alarm, and a coping action guide; and an output processing unit that outputs an event convergence display when an abnormality has converged. A monitoring means, a diagnostic means, an output processing means, and a control means are configured to operate independently of each other, and the monitoring means is activated by the control means. The plant monitoring process is performed at a predetermined cycle until a stop command is issued from the control unit, and the diagnosis unit is activated by a command of the monitoring unit when an alarm is issued, performs the diagnosis process at a predetermined cycle, and outputs the output. The processing means, when an alarm occurs or an abnormal event develops,
It is configured to be activated by a command of the diagnostic means to display a diagnostic result, and to be activated by the command of the monitoring means to display a monitoring result when an abnormal event converges and when a device to be operated does not operate. It is characterized by that.

In addition to the above configuration, another plant alarm processing apparatus according to the present invention has a plurality of knowledge databases corresponding to each plant operation mode, and the monitoring means automatically determines the operation mode of the plant. The knowledge database is switched over to perform alarm processing corresponding to all operation modes of the plant.

[0014]

In the plant alarm processing apparatus of the present invention, the monitoring means, the diagnosing means, the output processing means, and the control means operate independently of each other, and the operator can operate / stop the entire apparatus via the control means. The operator can monitor the plant in a state where psychological preparation for the abnormality alarm of the plant is completed. As a result, the diagnosis result of the alarm displayed when the abnormal alarm is issued can be quickly understood, and appropriate countermeasures can be taken.

Further, according to the plant alarm processing apparatus of the present invention having a plurality of knowledge databases corresponding to rated output operation, plant safety stop after abnormal convergence, restart after stop, and normal start, the monitoring means is provided. Since the operating mode is judged from the control signal of the plant and the like and the state of the plant is monitored by switching to the corresponding knowledge database, the alarm processing can be performed in all the operating modes of the plant.

Thus, the operator can activate the plant alarm processing device through the control means when he feels it is necessary in all the operation modes, and can more surely deal with the abnormality of the plant.

Further, in the plant alarm processing apparatus of the present invention, the monitoring means is activated by a command of the control means, monitors the plant alarm and control signal at a predetermined cycle, and if the alarm is generated, the diagnostic means is activated. To start. Since the knowledge database stores information such as attributes for each alarm, judgment conditions, causes / results of occurrences, countermeasures, causal relations of alarms, etc., the diagnostic means refers to the knowledge database and the alarm occurrence status and alarm transitions. With this, it is possible to identify the cause of the alarm, the prediction of the result of the alarm, the path of the abnormality, and the countermeasure. This diagnosis result is displayed to the operator when the abnormal event is progressing, so that the operator can accurately understand the cause of the abnormal event and the countermeasure.

Further, in the plant alarm processing apparatus of the present invention, even if new alarms are continuously generated, the monitoring means can judge from the knowledge of the knowledge database that the abnormal event tends to be converged, and if the alarm is converged. The operator can be shown an indication of the convergence of abnormal events. Further, when a device that should operate during the occurrence of an abnormality does not operate, an abnormality due to non-operation can be detected from the alarm sequence diagram of the knowledge database, and this can be displayed to the operator.

As described above, in the plant alarm processing apparatus of the present invention, the apparatus is activated in response to the operator's request during all the operation modes to periodically monitor the plant, and an alarm is generated and an abnormality develops. If there is an abnormal event, the diagnostic result such as the cause / result will be displayed.If the error converges even if an alarm occurs, the error convergence will be displayed.If no alarm should occur, a non-operation alarm will be displayed. The burden on the operator is greatly reduced, and the plant can be controlled with higher reliability.

[0020]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 schematically shows the configuration of an embodiment of the plant alarm processing apparatus of the present invention. The plant alarm processing device 1 of this embodiment is for a nuclear power plant, and includes a monitoring unit 2, a diagnostic unit 3, an output processing unit 4, and
It is composed of a control means 5 and a knowledge database 6.

Monitoring means 2, diagnostic means 3 and output processing means 4
The control means 5 and the control means 5 are configured to perform processing independently of each other. The monitoring means 2 is activated by the control means 5.
While being stopped and operating, the plant 7 is monitored by taking in various signals for alarms and process control of the plant 7 online at regular intervals.

The diagnostic means 3 is activated by the monitoring means 2 when an abnormality occurs, and the abnormality diagnosis processing is executed at a predetermined cycle (usually an integral multiple of the monitoring processing cycle of the monitoring means 2) until the abnormal event is converged. I do.

The output processing means 4 is activated by the diagnostic means 3 and displays the diagnostic result of the diagnostic means 3 to the operator 8 while the abnormal event progresses. Also, the output processing means 4
Is displayed by the monitoring means 2 to the operator 8 even when the abnormal event is converged and when the alarm should not be issued, the monitor 8 activates the display indicating that the abnormal event has converged and the alarm does not operate. To do.

The control means 5 sets data for starting and stopping the monitoring means 2 in response to an input from the operator 8. here,
Setting the data means an operation of temporarily storing an instruction input while the monitoring means 2 is performing processing and executing a start / stop instruction when the processing is completed.

The knowledge database 6 includes knowledge about the attributes of the alarm such as the name of the plant alarm and the condition for determining an abnormality, the cause of the alarm, the result of the alarm generation, the countermeasure, and the alarm generation. For each alarm, the importance of the alarm evaluated according to the magnitude of the impact on the plant and the knowledge about the conditions for alarm generation corresponding to the plant safety protection system, the plant interlock mechanism and the physical causal relationship are stored together for each alarm. ing. Further, the knowledge database 6 stores the knowledge obtained by developing the causal relationship of each alarm in a network form.

The knowledge database 6 is composed of a plurality of knowledge databases corresponding to the rated output operation of the plant, the safe stop after the abnormal convergence, the restart after the stop, and the normal start, respectively, and are switched by the monitoring means 2. The switched database is referred to by the monitoring means 2, the diagnostic means 3, and the output processing means 4.

Next, the operation of each constituent means of the plant alarm processing apparatus 1 will be individually described in detail below.

FIG. 2 specifically illustrates the switching of the knowledge database. Knowledge database 6 of this embodiment
Is the knowledge database 6a corresponding to the "driving" mode
And the knowledge database 6b corresponding to the "start" mode
And the knowledge database 6 corresponding to the "fuel change" mode
c and the knowledge database 6d corresponding to the "stop" mode
And have.

The knowledge database for each operation mode further has a plurality of sub-databases so as to apply different logics (alarm sequences, etc.) depending on the operation stage.

The monitoring means 2 takes in an operation mode switch signal of the plant and a device state signal or a process signal which is an index of operation mode switching, detects the operation mode at that time, and switches to the corresponding knowledge database. The monitoring means 2 of this embodiment includes a knowledge database 6a in the "operation" mode, a knowledge database 6d in the "stop" mode, a knowledge database 6c in the "fuel change" mode, and a knowledge database 6b in the "start" mode in this order. Select the corresponding database by switching.

The knowledge database of each operation mode has a plurality of sub-databases inside, and these sub-databases are switched in a predetermined order in a time series, and each sub-database is switched.
Switching between databases is performed by satisfying a predetermined condition.

FIG. 2 shows switching of sub-databases of the knowledge database 6a in the "driving" mode. The knowledge database 6a in the "operation" mode has a sub-database including a rated time database 9, a safe stop time database 10, and a restart time database 11. Several types of abnormal event occurrence / convergence / restart patterns are prepared for the alarm / sequence diagram of the rated time database 9. When an abnormality occurs during rated operation, the rated time database 9 is applied, and when the abnormal event is converged, the predetermined safety stop database 10 is applied depending on whether or not the abnormality convergence condition is satisfied. The judgment of the abnormal convergence condition is preferably made by a heuristic judgment. When the abnormal event is completely converged and the system is restarted, it is similarly determined whether or not the restart condition is satisfied and the mode is switched to the restart time database 11.

3 and 4 show the structure of the monitoring means 2 and the flow of processing. The monitoring means 2 comprises a signal input / mode switching means 2a, an alarm data synthesizing means 2b, a sequence operation monitoring means 2c, and an abnormality occurrence determining means 2d.

The monitoring means 2 starts its operation by the control means 5 (step 100), and inputs the signal of the operation mode switch of the plant, the ON / OFF binary signal indicating the equipment state, and the process signal (step 110). , The operation mode is determined (step 120) and the knowledge database 6 is switched. The above operation is performed by the signal input / mode switching means 2a.

Next, plant alarms and equipment status ON / OFF
Process signals such as signals and pressure / flow rate are taken online, applied to the knowledge (rule) of the knowledge database 6 to determine whether there is normal / abnormal, and the state of those signals is saved as the alarm state data of the cycle together with the determination result. (Step 130). In the subsequent monitoring process cycle,
Read previously saved alarm status data (step 14
0) Furthermore, the latest alarm status data is read and the two are compared to create status change data (alarm data) (step 150). The above operation is performed by the alarm data synthesizing means 2b.

Next, the diagnostic means 3 determines the predictive alarm data based on the alarm data, and the sequence operation monitoring means 2c monitors the operation of the device based on the predicted alarm (sequence monitoring, step 160). ). If there is a non-operation alarm, the non-operation alarm data is passed and the output processing means 4 is activated (steps 180, 190).

When a new alarm is issued (step 20)
0), it is determined whether or not the diagnostic processing started before the previous cycle is completed (step 210), and if completed, the latest alarm data is delivered to the diagnostic means 3 (step 2).
20) and activate the diagnostic means 3 (step 23).
0), the current alarm state is saved as the latest alarm state data (step 240). When the diagnostic means 3 has not completed the diagnostic process before the previous cycle, the current alarm state data is stored as it is as the latest alarm state data (step 240). The above operation is performed by the abnormality occurrence determination means 2d.

The operations of steps 110 to 240 described above constitute one cycle of the monitoring process of the monitoring means 2, and when the stage indicated by the symbol C is reached, the process returns to the upstream of step 110 and the above monitoring process is repeated.

Next, the flow of alarm processing when an abnormality is occurring during rated output operation but no new alarm is issued will be described with reference to FIG.

When a new alarm is not detected in step 200 of FIG. 4, the abnormal event convergence determination process of FIG.

When the subsequent alarm is not detected after the alarm is issued, it is determined whether or not the operation mode is the rated output operation (step 250), and then the abnormal convergence condition is determined (step 260). When the rated output is in operation and an error is occurring (all alarms are not reset), but if no subsequent alarm is issued, the alarm status data is not updated and diagnostic processing is not started. . In this case, when the abnormal convergence condition is further satisfied, the knowledge database 6 is switched to the database 10 at the time of the plant safety stop after the abnormal convergence.

If the abnormal event has completely converged and all the alarms have been reset (step 270), it is determined that the abnormal event has converged (step 280), and the event convergence display data is transferred to the output processing means 4 (step 290). ), And the monitoring means 2 activates the output processing means 4 (step 30).
0) Display event convergence data. Next, diagnostic means 3
Initialize after storing the diagnostic result data used in (step 310), and further initialize after storing the alarm state data used in the monitoring means 2 (step 320), and switch the database to the database at the time of restart. Next, it is judged whether or not there is an end request from the operator based on the monitoring / diagnosis process end request data (step 330). If there is an end request, the process is ended.
It returns to the upstream of 10 and repeats a new monitoring process. The processing of steps 270 to 330 is performed by the abnormality occurrence determination means 2d.

The flow of the above-mentioned diagnosis processing is the same in operation modes other than the rated output operation.

Next, the processing flow of the diagnostic means 3 will be described with reference to FIG. As shown in FIG. 6, the diagnostic means 3 is
Diagnosis result reading means 3a, latest occurrence warning reading means 3b, diagnosis execution means 3c, output processing starting means 3d
And a diagnostic result storage means 3e and a processing end transmission means 3f, and each processing means performs processing in accordance with each processing stage.

The diagnostic means 3 is activated by the monitoring means 2, and when activated, the diagnostic result reading means 3a reads the previous diagnostic result (step 400). Next, the latest occurrence alarm reading means 3b reads the latest occurrence alarm data (step 410). Next, diagnostic execution means 3
According to c, the alarm data at the latest and previous diagnosis is applied to the knowledge rule of the knowledge database 6 to identify the alarm progress path (step 420), the cause alarm is determined (step 430), and the alarm of the next occurrence is predicted ( (Step 440)
Then, the corresponding action is determined (step 450).

When the above-mentioned diagnosis processing is completed, the output processing starting means 3d transfers the present diagnosis result to the output processing means 4 (step 460), and the output processing means 4 is started (step 470). Next, the diagnosis result storage means 3e saves the present diagnosis result (step 480), and the process end transmission means 3f sets the diagnosis process end data (step 490) to end the diagnosis process. When the diagnostic process end data is set, a diagnostic process end flag is set, and the monitoring means can detect the end of the diagnostic process.

In diagnosing the cause and effect of the alarm and the progress route, it is necessary to consider the alarm generating condition which differs depending on the alarm generating cause. That is, two alarms a1 and a
In the causal relation between the two, if the alarm a1 is generated by the sequence operation of the plant safety protection system or the plant interlock mechanism, the alarm a2 is generated with a probability of 100%, but the alarm a1 is generated by pressure / pressure. The alarm a2 may not always be generated due to a physical cause such as a flow rate.

If the alarm a2 must be generated with a probability of 100% due to the alarm a1, the alarm a
If the alarm a2 does not occur even after the lapse of a predetermined delay time t from the occurrence of 1, it must be determined that the sequence operation is abnormal. This delay time t is long or short depending on the cause.

The knowledge database 6 stores knowledge about the occurrence condition for each alarm occurrence cause for each alarm, and the diagnosis execution means 3c checks the presence or absence of the issued alarm a1 to the alarm a2 generated by the sequence operation. Is determined as a predictive warning and registered together with the delay time (step 440). After that, after the delay time has passed, the monitoring means 2 checks whether or not a predictive warning has occurred,
If not, the output processing means 4 is activated to display a non-operation alarm.

Next, the flow of output processing of the output processing means 4 will be described with reference to FIG. As shown in FIG. 7, the output processing means 4 is activated by both the monitoring means 2 and the diagnosing means 3, and the display data reading means 4a reads the display data from the monitoring means 2 and the diagnosing means 3 (step 500). Next, the data type determining means 4b determines the type of display data, and the data displaying means 4c displays the non-operation alarm when the non-operation alarm is issued (step 52).
0), when an event progresses, an alarm progress route, a cause alarm, a prediction alarm, and a corresponding action guide are displayed (step 530), and when an event converges, the event convergence is displayed (step 54).
0).

Since the display of the non-operation alarm of the output processing means 4, the display of the diagnosis result such as the alarm progress route, and the display of the event convergence can be processed independently, the flow of the processing of the output processing means 4 is shown in FIG. It may be configured as follows.

According to the processing flow of the output processing means 4 of FIG. 8, the monitoring means 2 directly activates the display processing of the non-operation display and the display processing of the event convergence, and when each display processing is activated, the display data is displayed. The display data is read by the reading means 4a (steps 550, 570, 590), and the data is displayed by the data display means 4c (step 56).
0,580,600). According to this configuration, since it is not necessary to determine the type of display data, the data type determining means 4b can be omitted.

FIG. 9 shows a processing flow of the control means 5. The control unit 5 includes a process starting unit 5a and a process ending unit 5b.
The operator 8 is activated by a processing start command or the like, the operator 8 inputs start instruction data (step 700), and the monitoring means 2 is activated by the input start instruction data (step 710). When terminating the plant alarm processing apparatus 1, the operator 8 inputs termination request data (step 720) and requests termination of the processing of the monitoring means 2 by this termination request data (step 730). In this case, when the termination request data is input during the period in which the monitoring means 2 is performing the monitoring process, the termination request data is temporarily set in a predetermined memory, and the monitoring means 2
It is read when the processing cycle of is finished and the processing is finished.

As described above, according to the plant alarm processing apparatus 1 of the present embodiment, the monitoring means 2 is activated by the command of the operator 8 and is switched to the knowledge database 6 corresponding to the operation mode of the plant. The state of the plant is monitored according to the knowledge rule of 1., when an abnormality occurs, the diagnosis means 3 diagnoses the cause and result of the abnormal event, the progress route and the corresponding action, and the output processing means 4 displays the diagnosis result to the operator 8. Therefore, it is possible to support the judgment of the operator 8 in all operation modes, reduce the burden on the operator 8 and prevent erroneous judgments.

In the plant alarm processing system of this embodiment, the processing is started and ended when required by the operator 8 so that the operator 8 can monitor the plant with sufficient psychological preparation. In addition to the fact that the plant alarm processing can be performed in all of the above operation modes, highly reliable plant monitoring can be performed.

[0056]

As is apparent from the above description, in the plant alarm processing apparatus of the present invention, the operator activates the monitoring means by the control means, and the monitoring means monitors the plant at a predetermined cycle and an abnormality occurs. Occasionally, the monitoring means activates the diagnostic means, and the operator displays the diagnostic results such as the cause of the alarm, the predicted alarm, the path of the abnormal event, and the countermeasures to the operator through the output processing means. The plant can be monitored while being fully aware of the operation of. Further, the plant alarm processing device of the present invention has a knowledge database that stores alarm attributes and the like for each alarm and also stores a causal relationship between alarms. When an abnormality occurs, knowledge of this knowledge database is stored. Based on the above, the cause of the abnormal event, the result prediction, the abnormal event progress route, and the countermeasure can be displayed by the diagnostic means, so that the operator accurately grasps the plant state from the display of the alarm whose situation is difficult to judge, Appropriate action can be taken. As a result, it is possible to monitor the plant with high reliability.

In particular, the plant alarm processing apparatus of the present invention, which has a plurality of knowledge databases corresponding to the respective operation modes of the plant and is switched to the knowledge database corresponding to the operation modes by the monitoring means, is a plant alarm processing apparatus in all operation modes. Since it can monitor the alarm of, the operation load of the operator can be reduced, the judgment can be prevented, and the plant can be monitored with higher reliability even in the operation mode of the plant where abnormal events such as start-up operation and stop operation are likely to occur. It can be carried out.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration and processing flow of a plant alarm processing apparatus of the present invention.

FIG. 2 is a diagram exemplifying switching of knowledge databases of the plant alarm processing device of the present invention and switching of sub-databases of each knowledge database.

FIG. 3 is a flowchart showing a part of the processing flow of the monitoring means of the plant alarm processing apparatus of the present invention.

FIG. 4 is a flowchart showing a part of the processing flow of the monitoring means of the plant alarm processing apparatus of the present invention.

FIG. 5 is a flowchart showing a processing flow of a monitoring unit when a subsequent alarm is not detected during the occurrence of an abnormal event.

FIG. 6 is a flowchart showing a processing flow of a diagnostic means of the plant alarm processing apparatus of the present invention.

FIG. 7 is a flowchart showing an example of the configuration and processing flow of the output processing means of the plant alarm processing apparatus of the present invention.

FIG. 8 is a flowchart showing another example of the configuration and the flow of processing of the output processing means of the plant alarm processing apparatus of the present invention.

FIG. 9 is a flowchart showing the flow of processing of the control means of the plant alarm processing apparatus of the present invention.

[Explanation of symbols]

 1 Plant Alarm Processing Device 2 Monitoring Means 3 Diagnostic Means 4 Output Processing Means 5 Control Means 6 Knowledge Database 9 Rated Time Database 10 Safe Stop Database 11 Restart Database

Claims (2)

[Claims] 1. An alarm attribute including an alarm name and a judgment condition, an alarm generation cause, an alarm generation result, a countermeasure for the alarm, an importance of the alarm evaluated by an influence on a plant, and a generation cause. The knowledge database that stores the knowledge about the alarm occurrence conditions for A monitoring unit that detects the occurrence of an abnormality and creates alarm status data including the alarm occurrence status and the time of occurrence, and identifies the progression path of the abnormality from the alarm status data to identify the cause of the abnormality and the predicted alarm that will occur next. And a countermeasure for determining the corresponding action guide, a non-operation alarm when the device that should operate does not operate, and the alarm progress when the alarm occurs. A monitoring means and a diagnosing means, each of which includes a route, a cause alarm, a predictive alarm, a countermeasure guide, an output processing unit that outputs an event convergence display when an abnormality converges, and a control unit that sets start and end data. The output processing means and the control means are configured to operate independently of each other, the monitoring means is activated by the control means, and performs the monitoring processing of the plant at a predetermined cycle until there is a stop command from the control means, The diagnosis means is activated by an instruction of the monitoring means when an alarm is issued and performs a diagnosis process in a predetermined cycle, and the output processing means issues an instruction of the diagnosis means when an alarm occurs or an abnormal event develops. To display the diagnostic result, and when the abnormal event converges and the device that should operate does not operate, it operates by the command of the monitoring means and displays the monitoring result. A plant alarm processing device, characterized in that it is configured to display. 2. A plurality of knowledge databases respectively corresponding to plant operation modes are provided, and the monitoring means automatically determines the operation mode of the plant, and the knowledge databases are switched to correspond to each operation mode of the plant. The plant alarm processing device according to claim 1, which performs alarm processing.