JPH02245698A - Assisting device of plant abnormality diagnosis - Google Patents

Abstract

PURPOSE:To make a process of an abnormality diagnosis by an operator, much easier, and to enable an assistance of the abnormality diagnosis with a wide ranged realization and applicability by providing both an on-line input measure and a manual input measure of plant process informations. CONSTITUTION:A process data from a plant 11 is fed to a data memory 13 of an assisting device for plant abnormality diagnosis, by a data collector 12, and thereafter is output to a diagnosis device 14. To the diagnosis device 14, diagnostic logic informations from a diagnostic logic memory 15 which stores a previously prepared course consequence tree (CCT), and informations from a consequence memory 16 which stores a detailed abnormality consequence without a limitation by an input signal, are input, and then an abnormality of the process data is identified by the diagnosis device 14 and finally a diagnostic result is output to a display 17 of diagnostic informations. Also, input informations from an input device 18, after being input to a dialogue device 19 to interpretate the input informations, are input to an assisting device 20 of a diagnosis monitor, with which the abnormality diagnosis proceeds utilizing the diagnostic result of the device 14 and input informations from operators.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Industrial Application Field) The present invention relates to a plant abnormality diagnosis support device that diagnoses abnormalities in a plant using a computer system.

(Prior Art) Conventionally, CCT (Cause-Consequence) has been used as a device for diagnosing plant abnormalities using the causal relationship of abnormal events in plant equipment such as nuclear power plants.
There is a method using ence-Tree). This method uses logical operations to match the diagnosis logic that expresses the cause-and-effect relationships of abnormalities in the form of a logical tree and plant process data that is input in real time. It is used to diagnose the shadow tube and plant status.

FIG. 4 is a diagram showing the configuration of a plant abnormality diagnosis device using such a method.

Process data from plant 1 is collected by data collection device 2
The data is inputted into the data storage device 3 of the abnormality diagnosis device by , and then inputted into the diagnosis device 4 .

The diagnostic logic storage device 5 of the abnormality diagnostic device stores OCT that has been created in advance, and the input process data is compared with the diagnostic logic information from the diagnostic logic storage device 5 in the diagnostic device 5 to detect an abnormality. Identification is performed, and the diagnostic result is output to the diagnostic information display device 6.

In abnormality identification, in order to match the plant status described in each node of the CCT with the input process data,
Input data is processed. For the data to be compared, in the case of analog values, the result of determining whether or not the threshold value has been exceeded is used by the limit check method, and in the case of digital values, the results of determining values and changes are used. In CCT, logical operations are performed using these data, and abnormalities are identified from the established state of each node.

(Issue 8 to be Solved by the Invention) However, in the conventional abnormality diagnosis device described above, it is a premise that the process state expressed by each node on the diagnostic logic tree is input to the diagnosis device in real time. Therefore, in order to perform detailed diagnosis, it is necessary to input a huge amount of process data in real time, which makes it difficult to realize the system from an economic point of view. Further, since process data that is not directly coupled to a diagnostic device, such as a process without a sensor, cannot be used for diagnosis, the scope of application of diagnosis is limited.

Furthermore, based on the detection of an abnormality4, if we consider that the operator identifies the cause of the abnormality while performing verification on his own,
The operator must assume the cause of the abnormality based on the detected abnormal condition and carry out the plan 1-1 to verify the cause.

This process is based on the knowledge and experience of the operator, and omissions and duplication may occur during the verification process. Therefore,
Systematic support for abnormality identification and verification using detailed knowledge based on plant design is required.

The present invention was made in view of the above-mentioned conventional circumstances, and it compensates for the limitations of input process data with operator input, provides information necessary for operators to diagnose abnormalities, and It is an object of the present invention to provide a plant abnormality diagnosis support device that is capable of providing practical and widely applicable abnormality diagnosis support by facilitating the abnormality diagnosis process for personnel.

[Structure of the Invention] (Means for Solving the Problems) The plant abnormality diagnosis support device of the present invention comprises: an input means for inputting plant process information online; a storage means for storing the manually-generated process information; an abnormality diagnosis means for diagnosing abnormalities in a plant by performing logical operations based on process information; a causal relationship storage means for storing causal relationships of abnormalities in the plant; and a means for manually inputting estimated abnormality diagnosis information for the plant. a manual input means, the abnormality diagnosis result obtained based on the process information inputted online and the abnormality diagnosis information inputted by the manual input means are input to the abnormal event stored in the causal relationship storage means. This feature is characterized in that it is configured to support abnormality diagnosis by checking against causal relationship information.

(Function) Supports abnormality diagnosis by comparing diagnostic results from online information and abnormality diagnosis results input by operators with a knowledge base that shows detailed causal relationships of abnormalities that do not depend on online manual signals. Since this is done, it is possible to provide abnormality diagnosis support with a wide range of applications.

(Example) Hereinafter, an example of the present invention will be described with reference to the drawings. Note that this embodiment uses a knowledge base that describes design-based abnormality causal relationships that are not limited to input signals.

FIG. 1 is a diagram showing the overall configuration of the embodiment, in which a plant 11
After inputting the process data from the data collection device 12 to the data storage device 13 of the abnormality diagnosis support device,
It is output to the diagnostic device 14.

This diagnostic device 14 receives diagnostic logic information from a diagnostic logic storage device 15 that stores CCTs created in advance, and information from a causal relationship storage device 16 that stores detailed causal relationships of abnormalities not limited to input signals. is input, the diagnostic device 14 identifies abnormalities in the process data, and outputs the diagnostic results to the diagnostic information display device 17.

On the other hand, an input device 1-8 is provided for inputting operator's requests and diagnosis results detected by the operator, and input information input from the input device 18 is transferred to a dialog device 19 that interprets the input information. After the information is input to the system, the information is manually input to the diagnostic monitoring support device 20, which proceeds with the abnormality diagnosis using the diagnostic results of the diagnostic device and input information from the operator.

This diagnostic monitoring support device 20 receives information from a support logical storage device 21 that stores verification methods, monitoring items, etc. used in the diagnosis process, and information from the causal relationship storage device 16 and the diagnostic device 1t14. It is configured to

Incidentally, the causal relationship stored in the causal relationship storage device 16 may have a tree structure such as OCT, or a network structure indicating fault spread.

FIG. 2 is a diagram showing an example of the structure of the cause-and-effect relationship of such an abnormality, and in this example, a part of the abnormality of the reactor water supply control system in the nuclear couplant is shown in a tree structure. In the figure, a is O
R gate, b indicates AND gate, codes 31 to 48
010, Δ shown in is a node representing the plant state,
0 is the observable quantity that is directly input to the system, mouth is the observable quantity that is not directly input, and Δ is the unobservable ill ffi
It shows. An example of the node shape and g amount of each of these nodes is shown in the table of FIG. Note that an observable node may represent not only one process amount but also the logical operation results of multiple process ports. Furthermore, an unobserved node is automatically established when a lower logic gate is established.

For example, since the lower part of the node 38 is an OR gate, if any of the nodes 41 to 44 is established, the state is automatically established.

Plant abnormality diagnosis by the plant abnormality diagnosis support device having such a configuration can be divided into two processes. One is that the system automatically performs the diagnosis using signals input online, and the diagnostic device 14 performs this diagnosis.

The diagnostic device 14 narrows down the cause of the abnormality from the input signal. To narrow down the cause of the abnormality, search is performed downstream from the highest node where the cause is established. For example, in Figure 2, node 3
Assume that 2 is the top node. At this time, node 32
Since the lower logic gate is an O/R gate, it is sufficient that either node 33 or 34 is established. Therefore,
Check the status of nodes 33 and 34. The order of this search is the established input observable node, the non-input observable node,
The order is the input observable node that does not hold. For example, if node 33 is not established and node 34 is established,
The downstream side of node 34 is similarly searched. Note that in the case of a non-viewing node, the state is determined from a lower node state. When the search reaches node 37, its lower nodes 38~
40 are all non-observable nodes, and in this case, the lower nodes 41 and subsequent nodes are investigated and the established nodes are selected. Note that if there are no established nodes after node 41, the search is terminated at node 37, and 37 becomes the lowest established node.

Here, if the lowest established node is the abnormality cause node (nodes 41 to 44 and 48 in FIG. 2 correspond), the abnormality diagnosis ends, but if it is not the abnormality cause node, the diagnosis result is used as the diagnosis monitoring support. The data is sent to the device 20 and a second abnormality diagnosis is performed.

The diagnostic monitoring support device 20 identifies the abnormality while interacting with the operator using the result of narrowing down the cause of the abnormality by the diagnostic device 14, the abnormality causal relationship memory, and the device 16.

That is, since the abnormal conditions have been narrowed down by the diagnostic device 14 and the causes of the abnormal conditions are described by the knowledge base of causal relationships, the operator can verify the extracted abnormal causes for the abnormal conditions. good. At this time,
Anomaly causes for which failure has already been detected by the diagnostic device 14 can be excluded from verification targets. In addition, if the narrowed-down abnormal condition is still in the upper level state and is further expanded to non-human-powered non-abnormal cause nodes, the abnormal state is determined while asking the operator for human power in the state of the non-human-powered visible δ−j node. We will proceed with narrowing down the list.

In the above-described diagnostic process, abnormality diagnosis is performed using an abnormal condition detected by the diagnostic system as a trigger, but an abnormal condition detected by an operator can also be used as a trigger to request abnormality diagnosis support. In other words, when only a non-input observable state is detected by an operator, by inputting that state, a causal relationship from that state is searched, candidates for the cause of the abnormality are extracted, and verification methods and monitoring methods are developed. Information such as items can be provided.

Incidentally, in the above-described embodiment, the search order is shown unambiguously, but the priority can also be determined by setting the reliability based on the weighting of each node and the establishment state of the bus.

[Effects of the Invention] As explained above, according to the plant abnormality diagnosis support device of the present invention, not only the process values directly input to the diagnosis device (but also the operator's diagnosis results can be used), it is economical. It has excellent properties and can perform a wide range of diagnostics.

Further, since the diagnostic device or the operator can extract candidates for abnormal causes and verification methods from the diagnostic results, maintenance efficiency can be improved.

[Brief explanation of drawings]

FIG. 1 is a diagram showing the configuration of a plant abnormality diagnosis device according to an embodiment of the present invention, FIG. 2 is a logic tree showing a knowledge base of abnormality causality in the embodiment, and FIG. 3 is a diagram showing the abnormality causality in FIG. FIG. 4, which is a diagram for explaining the knowledge base of relationships, is a diagram showing the configuration of a conventional plant-level diagnosis device. 12... Data collection device 13...
...Data storage device 14...Diagnostic device 15...Diagnostic logic storage device 16...
...... Causal relationship storage device 17 ...... Diagnostic information display device 18 ...... Input device 19 ...... Dialogue device 20 ... ......Diagnosis monitoring support device 21...
...Support logical storage device applicant Japan Atomic Energy Corporation Applicant Toshiba Corporation Representative Patent attorney Sasu Suyama - Figure 1 Figure 2 Figure 4

Claims (1)

[Scope of Claims] Input means for inputting plant process information online, storage means for storing the input process information, and abnormality diagnosis means for diagnosing plant abnormalities by performing logical operations based on the process information. and a causal relationship storage means for storing causal relationships of abnormalities in the plant, and manual input means for manually inputting estimated abnormality diagnosis information of the plant, based on the process information input online. The system is configured to perform abnormality diagnosis support by comparing the obtained abnormality diagnosis result and the abnormality diagnosis information input by the manual input means with the causal relationship information of the abnormal event stored in the causal relationship storage means. Features: Plant abnormality diagnosis support device.