JPH0392799A - Abnormal-time operation support system of plant - Google Patents

Abstract

PURPOSE:To lower the possibility of misoperation to improve profitability by determining all actions together with an action for a caused event and having a corrective action determination part for selecting all the actions with respect to plant operation. CONSTITUTION:In a corrective action determination part 7 corrective actions on the basis of a knowledge data base 4 according to a sequence abnormality judged in a state monitoring part 3 and an important alarm selected in an alarm processing part is wholly determined. Here in the case where its judgement has been performed in a caused event judgement part 6, the action for a caused event and the other truly necessary actions on plant operation are determined and in the case where its judgement has not been performed, all actions with respect to the plant operation are selected as irreducible minimum information. Thereby since the corrective action and the other action on the truly necessary plant operation can be determined, a state judgement of operators increases certainty to lower the possibility of misoperation.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Field of Application in Industry) The present invention relates to an operation support system in the event of an abnormality in a large-scale plant such as a nuclear power plant, a thermal power plant, or a chemical plant.

(Prior Art) When an abnormality occurs in a large-scale plant, its influence is often observed as transient changes in multiple signals, and if the changes are large, multiple alarms are generated. Plan 1
・Operators are required to quickly judge the status of the plant from these multiple alarms and take appropriate measures, but if the magnitude of the abnormality is large or the spread of transient phenomena is rapid, , the burden on the operator becomes excessive, and
+'l'+I +There is a possibility of causing damage. For this reason, an alarm processing system is being developed with the aim of assisting operators in determining the situation in the event of an abnormality.

As such alarm processing systems, systems have been developed that determine the cause of an abnormality and the latest alarm, and predict future alarms. There was no event judgment function. Even if a cause alarm is determined by a conventional alarm processing system, subsequent response operations may differ depending on the initiating event. Therefore, operators must rely on their knowledge of plant structure and design and their own experience. Based on this knowledge, the alarm processing system calculates the result from the subsequent event progress.
1:. After confirming whether the error is correct or not, the cause of the abnormality must be determined and an appropriate treatment must be selected, leaving the possibility of misjudgment. In addition, some conventional alarm processing systems have been developed in which the predetermined order of alarm occurrence is stored as a database for each abnormal event, and the cause of the abnormality is determined by comparing the sequence of alarms that occur with the database. However, it is difficult to anticipate all possible events, making it impossible to respond when unexpected events occur, and the cost of creating and managing the database is high.

(Problems to be Solved by the Invention) The present invention eliminates the drawbacks of the above-mentioned prior art, and when an abnormality occurs in a large-scale plant, determines the cause of the abnormality alarm, selects important alarms related to treatment, monitors sequence operations, and issues cause alarms. The purpose of this system is to provide an abnormality driving support system that determines the initiating event that caused the occurrence, comprehensively determines countermeasures based on these results, and displays the results to support operators in making judgments about the situation. It is something.

[Structure of the Invention] (Means and Effects for Solving the Problems) In order to achieve the above object, the present invention provides information on the origin of the occurrence of each alarm in a plant, the occurrence result, and the influence that the occurrence of the alarm has on the operation of the plant. Conditions for determining the importance of the size of the feast and the presence or absence of an abnormality,
Safety protection system and Plan 1 determined at the time of designing Plan 1
・・Knowledge of plant operation measures for a series of sequence operations by interlock machines, and knowledge describing the relationship between alarms, alarms, equipment operating states, and process signals; In order to facilitate confirmation even when status confirmation is left to the driver, we have created a knowledge database that includes knowledge about what should be confirmed, its normal status, location, confirmation procedures, etc., and knowledge about measures to be taken in response to the vehicle phenomenon that caused it. By searching for knowledge about the cause of occurrence, importance, etc. for every alarm that occurs, and connecting the progress path of the alarm in a network, it is possible to determine the cause of the alarm and identify important alarms. After selection and sequence operation monitoring, knowledge describing the relationship between the triggering event and the alarm, equipment, and process signal for the cause alarm obtained after performing selection and sequence operation monitoring, and knowledge about the occurrence history of the event that is updated every time diagnosis is completed. If all the information necessary for determining the initiating event is obtained, the system can be used to determine the initiating event online, and determining the treatment for the initiating event and the truly necessary plant operation actions. Even if the necessary information is insufficient, use the minimum necessary information as a plan] After selecting all actions related to the operation of the plan, check the details of all possible events, their normal status, location, confirmation procedure, Outputs information in the form of a check list 1 for each event that suggests the order of confirmation so that the operator or himself can determine the initiating event and determine the corresponding action based on his knowledge of the actions to be taken against the initiating event. This feature is characterized in that it supports judgment of the situation of driving i-1 at the time of abnormality.

Therefore, according to the present invention, it is possible to determine the triggering event of the cause alarm, and to determine the corresponding action and truly necessary bran 1/operational action, so that the operator can more accurately judge the situation. This increases the possibility of erroneous operation.

(Example) Embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 1 shows a functional diagram of an embodiment of the present invention.
The abnormal driving support device 1 of the present invention is composed of a signal input section 2, a state monitoring section 3, a knowledge database 4, an alarm processing section 5, an initiating event determination section 6, a countermeasure determination section 7, and an input/output processing section 8. ing.

This device is activated at regular time intervals, and plant alarm signals, equipment status signals, and process signals are inputted from the signal input section 2.

The condition monitoring unit 3 uses the knowledge of the importance of each signal registered in the knowledge database 4 and the conditions for determining the presence or absence of an abnormality to monitor the plant from input alarm signals, equipment status signals, and process signals. Check for abnormal occurrences,
The correctness of each signal is saved as plant signal data 9 together with the determination result.

Also, before the previous diagnosis cycle, the alarm processing unit 5
The system counts down the delay time of the alarm that has been triggered, and checks whether or not it has occurred after a predetermined period of time has elapsed. Here, if the alarm has not occurred as planned, it is determined that the sequence is abnormal, and data to be displayed in the input/output processing section 8, which will be described later, is set.

The alarm processing section 5 is activated when an abnormality is detected in the condition monitoring section 3. Here, based on the knowledge about the causes, results, and importance of alarms registered in the knowledge database 4, the progress path of alarms in the plant signal data is created in a network form, and the cause-and-effect relationship of the alarms is determined. . As a result, the most second-rate alarm on the route is determined to be the cause alarm, and the most important alarm on the route is selected. Here, the causal relationship between the two alarms al and a2 includes, for example, ``If coolant pump trip J(al) occurs, ``Reactor scram J
(a2) occurs, there is a 100% probability that the occurrence of alarm a1 is caused by the safety system or the normal sequence operation next to the plant interlock machine.
They are connected by a physical cause-and-effect relationship, such that even if alarm a1 "coolant flow rate low" occurs, alarm a2 "coolant flow rate low" does not always occur. If there is, there are two types; the former type is an alarm
If the alarm a2 is not generated even after a predetermined delay time t has elapsed after the occurrence of the alarm a2, a sequence abnormality occurs. Therefore, we register which type is in the knowledge database 4, and if it is due to a sequence operation, we also register the delay time, and check whether there is an alarm a2 generated by the sequence operation from the lowest downstream alarm on the route. If there is, it is determined to be a predictive warning and registered along with the delay time.

Then, as described above, from the next diagnostic cycle onward, the condition monitoring section 3 checks whether a predictive alarm has occurred after a predetermined period of time has elapsed.

Knowledge database 4 includes the cause and result of each alarm in the plant, the importance level and abnormality judgment conditions that indicate the magnitude of the impact that the occurrence of the alarm has on the operation of the plant, and a series of information determined during the design of Bran 1. Knowledge of plan 1 and operational measures for sequence operations, knowledge that describes the relationship between the triggering event of an alarm, the operating status of the alarm and equipment, and process signals, and confirmation of these statuses is left to the operator. Even if
Knowledge regarding the content to be confirmed, its normal state, location, confirmation procedure, etc., and knowledge regarding actions to be taken against the occurring event are stored so that the confirmation and determination of the initiating event can be easily performed.

The cause event determination unit 6 determines which of the plurality of events that are considered to be the cause of the determined cause alarm is the true cause, based on the plan 1 and signal information for each engineering IS event stored in the knowledge database 4. The determination is made by comparing the state with the plant signal data 9 obtained by the state monitoring section 3.

Check whether all the plant signal states for each situation have been obtained, and if so, use the prefecture's knowledge of determining the triggering event for each alarm, and check the cause alarm, the latest occurrence alarm, equipment status, etc. After determining the occurrence online from the process information and determining the corresponding treatment, the occurrence history is saved. In addition, if all of the above signal states are not obtained, select the equipment to be investigated and communicate about it. For example, items such as actions to be taken are output to the operator in the form of a check sheet as shown in Table 1.In this case, the operator sequentially checks the status of the equipment, etc., and identifies any abnormalities. (This is written so that the initiating event can be determined in the end by determining the initiating event and the appropriate treatment can be decided.The order of these checks is based on the operation of the plan that directly affects the treatment. The higher the value is set, the more it is used to judge important events, and the more frequently it occurs than the event occurrence history data described below, which does not directly affect the treatment. The ranking remains unchanged unless the plant configuration itself is changed and the treatment procedure is reviewed, and the latter may change depending on the clothing store data saved each time a diagnosis is completed.

The response decision unit 7 comprehensively determines the response measures based on the sequence abnormality determined by the condition monitoring unit 3 and the important alarm selected by the alarm processing unit 5, beyond the knowledge database 4. . Here, if the determination is made in initiating event determination H6, the treatment for the initiating event and the truly necessary plant operation measures are determined, and if not, the minimum necessary information is used to operate the plant. Select all measures related to.

The input/output processing unit 8 performs input operations such as displaying the cause event and countermeasures, outputting the check sheet 1 of the equipment to be investigated, displaying the diagnostic results, and inputting the occurrence wear data 10 regarding the event determined by the operator. Perform output processing.

Next, a case will be described in detail below, taking as an example a case in which it is determined as a "cooling house pump trip" alarm or an abnormality cause alarm in a nuclear power plant.

If some abnormality occurs in the plant and the "coolant pump trip" alarm is determined as the cause alarm, the initiating event will be an abnormality (ground fault) in the power supply system that is the power source for the pump stick and cylinder. Possible causes include pump motor overload, erroneous signals, etc. Among these, the event that directly affects the subsequent treatment is Pumpstinoch. Therefore, first, from the latest alarm from the signal input section 2, we obtain information about the power supply system alarm related to the pump's power source and the occurrence of the "large pump vibration" alarm, and from the signal input section 2, we also obtain information about the pump's rotation speed and flow rate. . If the power system alarm has occurred, and it is determined that the "Pump Vibration Large" alarm has occurred and the flow rate and rotational speed of the pump have been determined to be a rapid flow course 1 down, then the "Cooling Shrine Pump" It was determined online that the initiating event that caused the "1. Rip" alarm was a pump stick, and from among the measures taken at the time of pump 1. Rip, the truly necessary plant operation - 1. 2. Removal of residual heat through natural circulation. The pump will be inspected and disassembled after a safe stop as a countermeasure for the problem caused by the vehicle failure. In addition, the power supply system alarm has been issued, and the "Pump Vibration Large" alarm has not occurred, and the normal flow course l.
If it can be determined that the pump is down, no pump stay has occurred, so as a plant operation measure, perform a blank l/safety shutdown procedure that excludes measures related to only the pump stick from among the measures taken during pump l/ripping. instruct. Here, the 11 causes may be an abnormality in the power supply system that is the power source of the pump (ground fault), and an overload of the pump motor.
・Outputs. According to this guideline, the operator first checks to see if there is an overload on the pump motor. From the sheet, compare the indicated value of the bomb motor current recorder installed at the 55th panel of A22 to the site with the Tokyo operation value of 45A, and if it is abnormal, it is determined that the initiating event is before the bomb motor overload. and suggests inspection of the motor.

Also, if it is normal, there is a possibility that there is an abnormality in the pump power source power system, so please refer to the item l shown in the power power system
” are sequentially checked to determine whether or not a candidate event has occurred. When the initiating event is finally determined, the event history is entered and saved for use in the next diagnosis.

FIG. 2 is a flowchart showing the processing procedure of the operation of the driving support system of the present invention described in "2", and the functional configuration for performing the processing procedure is correspondingly described on the right side.

Since this overlaps with what has already been explained, it will be briefly explained below.

When this driving support system is started, the first step 10
A plant alarm signal is input as 1.

In the second step 102, it is determined whether or not a predictive alarm has occurred, and if it is "absent", the display data of the alarm error, 71i, is set, and if it is "present", the third step 1
At step 03, it is determined whether a new alarm has occurred.

If a new alarm has occurred, in the fourth step 104, the cause alarm is determined and important alarms are selected, and then the next fifth alarm is generated.
In step 105, the next alarm is predicted. In the sixth step 106, the initiating event is determined, and in the seventh step 107, it is determined whether or not the initiating event has been determined.If rNOJ, the eighth step 10
In step 109, the diagnostic results are displayed. 10th step 1
In step 10, after outputting a check sheet, in step 11, the operator inputs the event history.

Further, if the initiating event has been determined in the seventh step 107, the event occurrence history is saved online in the next twelfth step 112. In the 13th step 113, a decision is made regarding the cause and operational measures to be taken for the cause small phenomenon, and in the next 14th step 114, the diagnosis results are displayed and the process ends.

In the above flowchart, the reason why the initiating event is determined even if no new alarm has occurred is that in the example of 2, the number of revolutions of the pump after t seconds have elapsed since the occurrence of the "coolant pump trip" alarm. This is because there are some cases, such as determining whether or not it is a normal flow coast down using γ and flow ix, which cannot be determined completely within one diagnostic cycle and will be performed after the next cycle.

Therefore, determination of initiating events is performed until all determinations are made, regardless of whether the latest alarm has occurred.

``Effects of the Invention'' As explained above, the abnormal situation support system of the present invention determines the triggering event of the cause alarm, and determines the corresponding action and truly necessary plan 1/operation 18. It is expected that this will increase the reliability of operators' judgment of the situation, reduce the possibility of erroneous operation, and improve economic efficiency by reducing the time spent on recovery. Ru. In addition, even when operators check the condition of equipment at the site to determine the cause, precise instructions allow them to concentrate on investigating the cause, reducing the burden on them. Since the system respects the autonomy of the operator by allowing the computer to operate the computer, it has the effect of eliminating human alienation caused by computers by υ1 and preventing human error.

[Brief explanation of drawings]

FIG. 1 is a functional block diagram of an embodiment of the present invention, and FIG. 2 is a flowchart showing the processing procedure of the operation of the present invention. ...Abnormal driving support system...Signal input section...Status monitoring section...Knowledge database...Alarm processing section 19 6...Cause event determination section 7...Response action determination section 8...Input/output processing Part 9, Plan 1, Signal data 10, Event occurrence history data (8733) Agent: Patent attorney Yoshiaki Inomata (and 1 other person) 20

Claims (1)

[Claims] The cause and result of each alarm in a large-scale plant such as a nuclear power plant, the degree of importance that indicates the impact that an alarm has on plant operation, the conditions for determining whether or not an abnormality has occurred, and the conditions determined at the time of plant design. Knowledge of plant operational measures for a series of sequential operations by safety protection systems and plant interlock mechanisms; knowledge of describing the relationship between initiating events that generate alarms and alarms, equipment operating conditions, and process signals; Knowledge of what should be checked, its normal status, location, confirmation procedures, precautions to be taken during confirmation, and measures to take in response to initiating events so that confirmation of these conditions can be easily performed even when it is left to the operator. A knowledge database that stores knowledge about the system, a signal input section that inputs plant signals such as alarms, equipment status signals, and process signals, and knowledge about the conditions for determining the degree of importance and whether an abnormality has occurred. a status monitoring unit that checks whether an abnormality has occurred in the plant from alarm signals, equipment status signals, and process signals, stores the value of each signal together with the determination result as plant signal data, and monitors sequence operations in the event of an abnormality; Alarm processing that determines the cause of the alarm, selects important alarms, and predicts the next alarm by connecting the progress paths of all alarms in a network based on knowledge about causes, results, and importance levels. Determine the initiating event based on the knowledge that describes the relationship between the initiating event and the alarm, equipment, and process signals for the obtained cause alarm, and the knowledge of the occurrence history of the event that is updated every time the diagnosis is completed. If all the information necessary for Determine the initiating event yourself,
If the initiating event determination unit creates information that even suggests the order of confirmation so that the treatment can be determined, and the initiating event determination unit makes the determination online,
Select plant operational actions from the results of alarm processing, extract truly necessary operations or monitoring items from the action items, determine comprehensive actions along with actions for initiating events, and if not. However, there is an input/output unit that selects all actions related to plant operation, outputs check sheets to operators, displays diagnostic results, and inputs occurrence history data of initiating events determined by operators. What is claimed is: 1. A plant abnormality operation support system comprising a processing unit and supporting an operator's situation judgment in an abnormality.