JPS59127105A - Display device for operation guide of plant - Google Patents

Abstract

PURPOSE:To improve both quality and reliability for operation guide of a plant through learning by showing an appropriate countermeasure to the plant based on the analysis result of known and expected phenomena when an accident arises at the plant and at the same time estimating an appropriate operation with an expected accident. CONSTITUTION:An accident model identifying part 2 fetches the primary process quantity of a main system, the changing factor of primary process quantity, the state quantity of system devices, the start signal and the start request signal from a plant 1 and then identifies an accident model if an accident sign is detected with variation of the primary process quantity. A learning part 4 adds a parameter of a new accident identification model and the information on an effective countermeasure to a guide data base 5 by making use of the time series data which are fetched and stored over the entire sequence section affected by an accident generated from a model parameter, i.e. a result of calculation obtained at the part 2. An operation guide deciding part 3 fetches the changing degree of the parameter obtained from a normal model of the identified accident model and decides an operation guide in response to the distance from each cluster in a parameter space of an expected accident model.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to a plant operation guide presentation device having a learning function that automatically presents operating instructions to an operator in the event of a nuclear couplant accident.

[Prior art]

In general, there are two types of systems for guiding plant operation. One method is to assume in advance an abnormality or accident that will occur in the plant, and to provide a ripple sequence of events in the form of tree logic based on the results of Ijh characteristic analysis, etc. Identifies all abnormalities or accidents. The driving guide is determined by selecting from the corresponding operations given to the assumed event sequence.

Other methods detect signs of an abnormality or accident from changes in the plant's main process quantities or changes in the state of system equipment, and then select countermeasures that will return the plant to normal conditions. This is a pace method.

By the way, the former method is inappropriate when an unexpected event occurs in the plant or when it is not possible to understand the occurring event in detail. On the other hand, the latter method does not require elucidation of the cause of the event, so it can be applied regardless of the presence or absence of an unexpected event, and is suitable for cases where safety is paramount, such as in the case of a plant accident. However, since response operations are selected only based on symptoms without determining whether an accident has occurred, there is a risk of unnecessary or excessive driving operations, and the knowledge of the analysis results of many hypothetical events cannot be reflected. Therefore, it is difficult to say that this is a high-quality driving guide method.

[Purpose of the invention]

The purpose of the present invention is to present an appropriate response operation for a plant in the event of a plant accident based on the analysis results of known assumed events, and also to be able to estimate appropriate response operations even in the event of an unexpected accident. An object of the present invention is to provide a plant operation guide presentation device that can improve the quality and reliability of the operation guide.

[Summary of the invention]

The present invention automatically processes data from a plant to identify an accident and presents an operator with an operation guide showing countermeasures for the accident. an accident model identification section that identifies the type of accident, a guide database that stores data on various assumed accident models prepared in advance, and data related to the identified accidents from the accident model identification section and data on the assumed accident models in the guide database. A driving guide determining unit that determines a driving guide indicating response operations for the identified accident and presents it to all operators, and imports data from the plant at the time of the accident and data regarding the identified accident from the accident model identification unit. The above objective is achieved by sincerely adopting a driving guide presentation device consisting of a learning section that adds new response operation information regarding known hypothetical accidents or response operation information regarding new accidents to the guide database. do.

Next, the principle of the present invention will be explained.

Learning in the plant operation guide function can be roughly divided into two types. The first is to add knowledge of accident events that occur in plants and effective operation methods to deal with them.
The other is a function that correctly corrects when the prior knowledge of the actual accident and response operations does not match the plant characteristics at the time of 5A. In the present invention, in order to realize this function, knowledge of accidents and response operations (hereinafter referred to as a guide database) is concisely expressed in a plant accident model and response operation table. Therefore, since plant accidents are caused by failures or structural damage to plant control system equipment, we focused on the fact that parameter variations in the plant accident model from the normal model are associated with each accident event.

In addition, it has a function to respond to unexpected events, which is a feature of the symptom pace driving guide method, and it also has detailed knowledge about expected events! ? For the purpose of improving the quality of driving guidance using the above-mentioned parameter space, it is possible to estimate the corresponding operation using the concept of distance in the above parameter space. This is because, in the parameter space, points that correspond to similar accident events are concentrated and form clusters, and the operating operations that should be taken at the time of an accident are limited to those related to controlling the main process quantities. This is because the same driving operations are thought to correspond to similar accident events that form clusters. It is thought that the unexpected event does not belong to any cluster, but in this case, the degree of similarity between events is defined by the distance to the center of gravity of each cluster, and this is reflected in the selection order of driving maneuvers to be taken or in determining the operational procedure. let

The present invention will be explained in detail below. FIG. 1 is an overall block diagram showing an embodiment of the plant operation guide presentation device of the present invention. A plant generally consists of a main system and subsystems, and in the case of a nuclear power plant, it consists of a nuclear reactor, a turbine, a generator, and system equipment such as a main control system, auxiliary system, and safety protection system. The accident model identification unit 2 takes in the main process quantities of the main system, the rate of change of the main process i, the state quantities of system equipment, the start signal, and the start request signal from the plant 1, and detects changes in the main process quantities of the plant as signs of an accident occurrence. If this is obtained, model identification is performed.

The learning section 4 takes in the chain state quantities of the plant and the model parameters that are the calculation results of the accident model identification section 2, and uses the time series data that has been taken in and stored for the entire spread sequence of the accident event that occurred. The parameters of the new accident identification model and information on effective response operations are added to the guide database 5.

Furthermore, while the accident event is spreading, the change rate of the main process quantities among the chain state quantities of the plant, the state quantities of the system equipment, the start signal, and the start request signal are used to calculate the change of the main process quantities due to the operation of the system equipment. The control effect is calculated, and if it differs from the data given to the guide database 5, the data is corrected.

The driving guide determining unit 3 takes in the parameter changes of the identified accident model from the normal model given in advance, determines points in the parameter space prepared in the guide database 5, and calculates each point in the parameter space of the assumed accident model. A driving guide is determined according to the distance from the cluster and displayed to the operator 6.

Next, a case where an embodiment of the present invention is applied to a boiling water nuclear power plant will be described using detailed configuration diagrams of each functional block. FIG. 2 is a functional block diagram showing detailed configurations of the accident model identification section 2 and the driving guide determination section 3. The main process amount variation calculation unit 15 is the main process lid measurement unit 1
0, the main process it is taken in from the plant 1 and outputs the main process amount fluctuation amount to the corresponding operation selection section 18. The plant structure model parameter change calculation unit 16 receives the main process quantity change rate from the plant 1 through the main process variable change rate measurement unit 11, and also inputs the main process quantity change rate from the plant 1 through the main process quantity change rate measurement unit 11.
2, the operation input is taken in from the plant 1, and the structural model parameter variation amount is output to the corresponding operation selection section 18.

The system equipment model parameter change calculation unit 17 takes in all the system equipment start signals from the plant 1 through the system equipment start measurement unit 13, and also receives the thread g equipment start request signal from the plant 1 through the system equipment start request signal measurement unit 14. The system equipment model meter fluctuation amount is taken in from the handling unit 1 and outputted to the corresponding operation selection unit 18.

The corresponding operation selection section 18 outputs a series of corresponding operation driving procedures to the driving guide display section 19 based on various input data, and the driving guide display section 19 displays this.

Next, the functional operation and arithmetic algorithm (9) of this embodiment will be explained. There are various methods for identifying a plant model, but in the embodiment of the present invention, a simplified input/output model is created based on a linear state equation, as shown below. Now, if the operation input due to the activation of the system equipment of the plant is expressed as δu1 and the variation in the rate of change of the main process amount at this time is expressed as a δ statement, it can be approximated as shown in the following equation.

δX=S·δU (1) Here, BFi is an input system matrix, and since this is determined by the dynamic characteristics specific to the plant main system, this element is called a plant structural model parameter. In addition, the operating human power δ
U sends the system equipment activation signal and the same request signal to δW, δwe
It can be expressed by the following equation.

δu=U・δW...(2) δw=
K・δW*...(3) Here, U
K is a diagonal matrix that shows the relationship between system equipment operation and control input; The following equation is obtained from equations (2) and (3).

(10) δu=C・δW4...(4)
Cmit J-K...(5
) Here, since the matrix C is determined by the functional state specific to the system equipment, this element is referred to as the entire system equipment model parameter. In the event of a plant accident, it is necessary to select all operating procedures that prioritize safety. For this reason, the main process quantities will be those related to safety functions, and Table 1 shows the main process quantities and related system components for a boiling water nuclear power plant. Limit values are set for each major process quantity to ensure the safety of the reactor.

For example, for the reactor water level, the normal operating water level is 1157
+a+, the reactor scram water level is 273a+, and the automatic startup water level of the emergency core cooling system (eccs system) is set at -122m (high pressure system) and -3720m (low pressure system), which are the heights from the bottom of the steam-water separator. ing.

Therefore, if there is a fluctuation in the reactor water level and its value reaches the above-mentioned set value, it can be assumed that an accident will occur that will lower the reactor water level.

The main process quantity variation calculation unit 15 takes in the main bromo (11) quantity, compares it with the set value for variation evaluation, and determines that an accident has occurred when the set value is reached, and the response operation selection unit 18
Start. The accident model identification unit separates the plant structure model and the system equipment model and identifies deviations from the respective normal models, and the basic equations of the models are equations (1) and (4). Table 2 shows typical assumed accident events and changes in model parameters and corresponding operations? show. For example, when the water supply function is lost, water cannot be injected into the reactor in response to the water supply system operation request command, so the system equipment model parameter C
The element CfW related to the water supply function changes from 1 to 0. It is thought that water injection decreased during the loss of cooling water accident (LOCA). This results in elemental fluctuations in the plant structure and the chill meter. To identify each model parameter, calculations are performed by incorporating the attendance variables in equations (1) and (4).

The plant structure model parameter change calculation unit 16 takes in the six main process quantity change rates and the system equipment state quantities (12), Table 1 (13), and Table 2 (14), and calculates the plant structure using the following formula. Evaluate the element variation of model parameter B.

Assume that system device j starts up.

HI = 1 (2 sentences) Umbrella - 6 sentences 1 ... (6) (
j=1. ... m) (δ month = B4 δU ... (7) When the system equipment performance index given by equation (6) becomes larger than the threshold value, the given plant structure model parameter B* is corrected. Now, when the high-pressure core spray system is activated as a system equipment, the main fluctuation element of the operation input vector is the water injection term U.Therefore, the amount of fluctuation (δB) of the W-th column component of the plant structure model parameter B is as follows. It can be found by the formula.

1 ・ (δB ) , −−(δX −6 text) ・・
・(8) δu.

Similarly, the system equipment model parameter change calculation unit 17 takes in the operation input δU from among the system equipment start signal δW1, the system equipment start request compensation code δw” and the system equipment status,
The amount of variation in system equipment model parameters] δC) is determined.

For example, when the high pressure core (15) pre-system startup request signal (6w'') itpcg is output, it is determined by the following equation.

(−〇) *pcm=δU−δu”IPCI
-(9) However, δW''mpcm=1. If the high-pressure core spray system does not start due to an abnormality, the first term on the right side of equation (9) becomes δU=O.

Additionally, the functionality of system equipment generally has inherent nonlinear characteristics that change when major process quantities vary significantly. For example, as the reactor pressure increases, the pump discharge flow rate decreases. Since such variations in characteristics have no relation to the occurrence of an accident, the given system equipment model parameter C umbrella is successively corrected according to the function expression of the main process quantity.

Next, the operation of the corresponding operation selection section IB will be explained.

The guide database 5 includes parameters whose coordinate axis components are the elements of each parameter matrix, representing the variation of plant structural model parameters and system equipment model parameters obtained from the analysis results of assumed accident events using equations (8) and (9). It has four pieces of data that are clustered using the multivariate (16) analysis method, which is data displayed in space and a set of points concentrated in a certain spatial area. At the same time, it is assumed that corresponding operations are prepared for each cluster in the form of a table. For example, as shown in Table 2, in cases where similar model parameter fluctuations occur, such as loss of water supply function and loss of off-site power accidents, countermeasures include using the isolation cooling system (CIC) and high-pressure core spray. System (H
The same corresponding operation is selected as required to start up the PC 8). Now, when changes in plant structure model parameters and changes in system equipment model parameters are given, a point P in the parameter space shown in FIG. 3 is determined.

The corresponding operation selection unit 18 first calculates the distance to the center of gravity of each cluster. The center of gravity k(,xIamXlt
e...Xl,), the Euclidean distance L% is expressed by the following formula.

Operation U corresponding to each cluster CJ (' = L...h)
When t ('=ts...h) is given, the synthetic response operation U to be selected for the accident event at point P is determined by the following equation.

1 and the combined corresponding operation U can be interpreted as the effect of each corresponding operation...the size of the weighting coefficient indicates the priority order of corresponding operation execution. Alternatively, the accident event that has occurred may be significantly different from the assumed accident, and different response operations may need to be carried out at the same time. Which interpretation to adopt depends on the plant situation and is at the operator's discretion. It is also possible to provide a series of operating procedures as corresponding operations for each cluster. In this case, the synthesis correspondence operation is obtained by performing a weighted addition of the reciprocal distance in each unit section when time is divided into unit sections.

That is, a series of corresponding operation driving procedures can be expressed in a vector format, and a composite corresponding operation can be obtained by addition using vector operations.

Next, the operation of the learning section 4 will be explained in accordance with FIG. The learning section 4 has two main functions: an offline driving guide learning section 25 and an adaptive driving guide modification section 26. In the offline driving guide learning unit 25, the main process quantity response storage unit 20, the accident identification model parameter change storage unit 21, and the system equipment operation sequence storage unit 22 store each input data during a series of accident event sequence periods in chronological order. Remember it. The main process amount time series data evaluation unit 23 has two functions. One is to evaluate the degree of fluctuation by successively comparing the main process quantities and the operational limit value while the accident event is in progress, and determine that the accident sequence has ended when the degree of fluctuation becomes smaller. The other method is performed after the accident sequence ends, and the stored time-series data of the main process quantities is retrieved for each unit time interval and the fluctuation tendency is evaluated. This evaluation obtains the degree of deviation of the main process quantity from the set target value in the form of an index, and also takes into account the rate of change, as shown in the following equation.

(t=1...n) Here, δXl+δX is the deviation between the main process quantity and the target value of its rate of change, and hsllhll is the standardization constant (19). The validity of the driving operation being carried out can be evaluated based on the fluctuation trend in each time period of the deviation index given by the round equation. The guide database generation unit 24 extracts accident model parameter changes and system equipment operation sequences from the stored data for each time segment, and evaluates the validity of corresponding operations by system equipment operation based on the value of the deviation index. As a result, if the validity of the response operation is confirmed as the deviation index shows a decreasing trend, this response operation is paired with the accident model parameter change data and added to the guide database. In this way, the experience of accidental events in the plant is learned.

On the other hand, the adaptive operation guide modification unit 26 takes in the main process quantity change rate, the system equipment state quantity, the system equipment activation signal, and the system equipment activation request signal, and evaluates the functional state of the system equipment. This evaluation involves determining whether or not a failure has occurred in the system equipment and comparing the value of the operation input with the target value in the rated state. Based on the results, the degree of effectiveness of each corresponding operation in the guide database 5 is corrected.

(20) FIG. 5 is a flowchart showing the operation timing of this embodiment. First, in step 101, 4'l1 is added to 'whether or not the main process quantity has changed, and if it has changed, a plant accident model is identified in step 102,
Next, in step 103, the parameter is set to the distance # between
The f calculation is performed, and in step 104, the corresponding operation is determined based on the guide database. Next, in step 105, it is determined whether the response of the main process quantity is as expected or not. If it is as expected, the process is terminated. If it is not as expected, the contents of the guide database are modified in step 106. and end the process. On the other hand, if the main process amount has not changed in step 101, it is determined in step 107 whether the accident sequence has ended or not. The relationship between the accident model and the response operation is determined in step 109, and this is added to the guide database in step 109, and the process ends.

According to this embodiment, the accident model identification unit 2 (21
) The system takes various state quantities from the plant at the time of the accident, calculates the parameter fluctuations from the normal model stored in advance, identifies the accident, and calculates the parameter fluctuations to the operation guide determining unit 3.
compares the characteristics of known accident event groups prepared in the imported guide database with the extracted data group (performed in the C parameter space), determines a driving guide for the accident, and displays this to the operator 6. In particular, it is possible to suggest appropriate plant response operations in the event of an accident based on the analysis results of known expected events, and even when an unexpected accident occurs, it is possible to provide a guide database that provides an approach to the group of known accident events prepared in the guide database. Then, by synthesizing and determining an appropriate driving guide and displaying it to the operator 6, appropriate response operations can be performed even when an unexpected accident occurs. Therefore, there is an effect of preventing excessive response operations or response operations that are too late, and allowing safe operation of Plan 1).

In addition, the learning unit 4 takes in the various state quantities of the plant and the parameter fluctuations from the accident model identification unit 2, and fully opens the time series data stored in the entire interval (22) of the spread sequence of the accident event that occurred. In addition to adding new accident identification model parameters and information on effective response operations to the guide database 5, during the propagation of the accident event 9, some data of the state quantities of the plant are used to update the system. The learning effect of calculating the control effect of the main process quantities due to equipment operation and correcting it if the calculation result differs from the data given in the guide database 5 increases the reliability and reliability of the presented operation guide. Quality can be constantly improved.

〔Effect of the invention〕

As described above, the plant operation guide presentation device of the present invention has the effect of presenting appropriate plant response operations in the event of an accident based on the results of analysis of known assumed events, and is also effective in presenting appropriate plant response operations in the event of an unexpected accident. It has the effect of estimating and presenting the necessary response operations, and also has a learning function that corrects or adds to the expected event analysis results based on data analysis at the time of each accident, improving the quality and reliability of the driving guide. It has the effect of

(73)

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of the plant driving guide presentation device of the present invention, FIG. Fig. 3 is an explanatory diagram showing an example of a cluster of accident models displayed in the parameter space, Fig. 4 is a block diagram showing an example of the detailed configuration of the learning section shown in Fig. 1, and Fig. 5 is an illustration of the plant of this embodiment. The operating timing of the driving guide presentation device. It is a flowchart figure shown. 1...Plant, 2...Accident model identification section, 3...
・Driving guide determination section, 4...Learning section, 5...Guide database (24) Eye 4 ◇ 2nd Eye 4 Eye Z5 26- Figure 5

Claims (1)

[Scope of Claims] 1. An accident model identification unit that identifies an accident model by calculating parameter fluctuations that are deviations from a normal model given in advance based on input main chain state quantities of the plant; , a guide database that holds a data group in a parameter space indicating the characteristics of an accident event group obtained from the analysis results of hypothetical accident events that are expected to occur in a plant, and a driving guide data group corresponding to each accident event group, and an accident model. A driving guide presentation device for a plant, comprising a driving guide determining unit that compares the amount of parameter variation from the same leg with a data group in a parameter space of a guide database and determines a driving guide to deal with the accident. . 2. Incorporate and analyze the chain state quantities from the plant at the time of the accident and the parameter fluctuation quantities from the accident model identification unit, and generate countermeasure operation information regarding known hypothetical accidents or countermeasure operation information regarding new accidents. 2. The plant operation guide presentation device according to claim 1, further comprising a learning section for adding the information to the guide database. 3. The driving guide determining unit determines a point in the parameter space stored in the guide database based on the input parameter variation, and calculates the distance of this point to the plurality of accident model clusters in the parameter space, The plant operation guide according to claim 1, characterized in that each operation guide corresponding to each accident model cluster is synthesized according to this distance to create a series of operation guides for dealing with the accident. Presentation device.