JPS63149729A - Device for supporting formation of knowledge base - Google Patents

Abstract

PURPOSE:To support the formation of knowledge bases necessary for the operation of a plant or the like by forming relation between events including operation and plant status from the operation recording data of the plant as knowledge data and graphics. CONSTITUTION:Data relating to the plant status (events) generated by the operation of the plant and its execution are read out from an input device 10, an event relating rule indicating the relation among events stored in a storage device 30 is retrieved about the read events and data (knowledge) indicating the relation among the events is formed and stored in a storage device 31. A graphic signal arraying the events in accordance with the existence of relation is found out on the basis of the knowledge and displayed on a display device 50. When an edition ending instruction is inputted from a mouse 40a or a keyboard 40b, the knowledge is stored in the storing means 31 to end the processing. When the edition ending instruction is not inputted, an input instruction for a correcting part of the displayed graphic or its correcting contents is discriminated to correct the knowledge indicating the relation among the events based on the input instruction.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a knowledge base creation support device, and particularly supports the creation and maintenance of a knowledge base used by a system that provides knowledge necessary for operating a plant, etc. This invention relates to a knowledge base creation support device.

[Conventional technology]

Conventionally, many of the devices that support the creation of knowledge bases support the analysis and confirmation of contradictions and omissions in existing knowledge, as described in Japanese Patent Application No. 59-92815. It did not support the creation itself. An example of automatic knowledge creation is ``Development of an automatic generation system for safety-related knowledge of atomic couplants'' 2 Atomic Energy Society of Japan.

As stated in the 1986 Society, E 14 (March 1986), design information is based on static data such as structural data regarding the connections and structures of each component and sensor, and physical property data regarding the properties of materials. To.

Some attempts are made to automatically derive abnormal symptoms through qualitative simulation. In other words, this method is characterized by performing a qualitative simulation using design information instead of a conventional quantitative simulation using a physical model. Compared to quantitative simulations, this method saves calculation costs and time, and is also characterized by the ability to automatically generate knowledge about abnormal symptoms.

[Problem that the invention seeks to solve]

In the above-mentioned conventional technology, the generated knowledge is naturally limited to that derived from design information, and the knowledge that is generated is limited to that derived from design information, and the knowledge that is generated is limited to that derived from design information. The derivation of knowledge such as procedures is not taken into consideration. Operating methods such as operating procedures are determined depending on information other than design information such as the purpose of operation, and the knowledge derived by the method described above is not considered. Automatic generation is difficult.

An object of the present invention is to provide a knowledge base creation support device that can support the creation of a knowledge base necessary for a system that provides knowledge necessary for operating a plant or the like.

[Means for solving problems]

The above objectives were prepared in advance based on information on plant conditions (hereinafter collectively referred to as events) such as operations and system/equipment operations, changes in state quantities, etc. caused by operations, such as operation record data in plants, etc. Create knowledge data about the interrelationships between events using rules that indicate the relationships between events.

Furthermore, based on this created knowledge data, a graphical signal in which events are arranged according to the presence or absence of a relationship is determined and displayed.
Alternatively, this can be accomplished by configuring the knowledge data and display graphics to be modified based on input instructions.

[Effect]

According to the present invention, from plant operation record data, etc.,
Relationships between events including operations and plant conditions are automatically created as knowledge data and graphics. Furthermore, it is easy to edit the data after creating 5 to make it easier to use.

〔Example〕

An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of the present invention,
Data input device 10. Arithmetic device 20. Storage device 30, 3
1. It is composed of command input devices such as a mouse 40a and a keyboard 40b, and a display device 150. Arithmetic device 2
0 includes, in addition to the arithmetic unit 20a, a data processing unit 2 that functions as an interface with various external devices.
0b, event-related rule search unit 20c, search result input unit 20
d.

There is a knowledge output section 20e, a knowledge input section 20f1, an image data output section 20g, and an input command processing section 20h. Also,
The memory device N30 is predetermined. Event-related rules specifying the relationships between events are stored in the storage device 31, while the storage device 31 stores data (hereinafter referred to as (referred to as knowledge) is stored.

FIG. 2 shows the calculation section v! As shown in Figure 0, which is a flowchart showing an overview of the borrowing process performed in step 120, in the arithmetic device 20, first, the input device! 10 regarding the plant operation and the plant state (event) caused by its implementation, is read through the data processing unit 20b (step 61
)1 Next, the event-related rules indicating the relationships between events stored in the storage device 30 are retrieved for the read events (step 62), and data (knowledge) representing the relationships between the events are created (step 63). The knowledge is necessary to create a signal for displaying on the display device 50 a figure in which events are arranged according to the presence or absence of a relationship.6 The created knowledge is stored in the storage device 31 (step 64),
Next, the previously stored knowledge is read from the storage device 31 (step 65), and based on that knowledge, a graphic signal in which events are arranged according to the presence or absence of a relationship is obtained (step 66).
.

The graphic signal is sent to the display device via the image data output section 20g! ! Display the figure by sending it to 50! ! Next, the mouse 40a or the keyboard 40
It is checked whether a command to end editing has been inputted from the command input device b (step 68).As a result, if a command to end editing has been inputted, the knowledge is stored in the storage device! 31 and ends the process (step 71). On the other hand, if the editing end command has not been input, the modified part of the displayed figure 1,
Step 69)
, modify the knowledge representing the relationship between events based on the input instructions (step 70).Then, the process from step 66 is repeated.

The input data to the device of this embodiment may be event-related data such as operation training records using a plant simulator or operation records at an actual plant.On the other hand, the storage device! The format of the knowledge stored in 31 does not need to be a specific format as long as it is possible to express the relationship between events; for example, it may be expressed in a 1f-then rule format, a frame format, or the like. Depending on the input data or the expression format of knowledge, the event-related rules in the storage device 130 or the display format of graphics on the display device, etc. will be somewhat specific. Here, the processing in the apparatus according to the present embodiment will be specifically explained using the following input data and knowledge expression formats as an example.

FIGS. 3 to 8 are diagrams for specifically explaining the processing in the apparatus according to the embodiment. Of these, FIG. 3 shows an outline of input data.

As shown in the figure, the input data used here is a chronological arrangement of events that occur in the plant, categorized into operations of nine operating devices, alarms, etc., and changes in plant state quantities. In Figure 0, 01g C1*P1 etc. is 1, for example "Test switch ON".

This indicates an event such as "Valve #1 open" or "Pressure rise."

4 and 5 are diagrams showing an outline of the common sense algorithm (hereinafter referred to as C8A) selected as the knowledge expression format. C8A is a method for expressing dynamic knowledge regarding the relationship between an action, a state, a state change, etc. Knowledge is expressed using links that indicate the relationships between events (hereinafter referred to as events) and the relationships between events. In addition, events and links include
Several types are available. As illustrated in Figure 4, the knowledge expressed by the CS system includes event-related data such as the knowledge name, event, event number, and event type, as well as the type of link and the events connected by the link (output and input). ), and for conditional links, the event that is the condition, and other link-related data.

1± in C5A, it is also possible to express it diagrammatically 1 For example, as shown in Figure 5, events such as O engineering t C1e Pl,
The content of knowledge can be expressed by the links that connect them. Here, the type of link.

It is expressed by changing its shape.

FIG. 6 is a flowchart showing the processing at the time of knowledge creation when using the above input data and knowledge expression format, and specifically shows the processing of steps 61 to 64 in FIG. 2. As shown in Figure 0.

When creating knowledge, first, the events included in the input data are read in order from the earliest occurrence time (step 601), and then the read events are stored in the storage device! Step 6: Search for event-related rules within 30.
02), determine the event type (step 803)
), an example of the event-related rules used for this determination is shown in Figure 7. Among these rules, the event type determination rules are as follows:
For example, if an event is classified as an operation, the event type may be determined as rAction J. Furthermore, there is a method for specifying the event N class for a specific event.

After determining the event type in step 603, a link type determination rule is searched (step 604).

Determine the type of link (step 605), where:
The rule for determining the type of link is that a link'! There are options for determining iR1[, or for specifying the type of link to a specific event.

Once the event and link types have been determined through the above steps,
Adding data such as the name of the knowledge, C5A format knowledge including the data shown in Figure 4 is created (step 606).
. Then, it is stored in the storage device 31 (step 6
07).

After the creation of the knowledge is completed, the knowledge is displayed graphically and the knowledge is modified in accordance with the input command. here,
Processing when both the mouse 40a and keyboard 40b are used as input devices will be described.

Figure 8 shows 1 display bag! ! 1 which is an example of the display screen at 50
On the screen, to make it easier to enter commands using the mouse,
I icons 81 to 83 are provided. In addition, to input a string from the keyboard, first enter the string input field lOO.
Move within the screen according to the movement to the 0 screen displayed on the screen.

For the I icon.

81 related to event modification, 8 related to link modification
2, and 83 related to general control. By moving the mouse cursor to each icon and pressing the mouse button, the process corresponding to each icon is executed.

Follow the steps to edit knowledge and provide an overview of the process. The processing here corresponds to step 65 and subsequent steps in FIG. When the mouse cursor is placed at the knowledge display position of Fitun 83 and the mouse button is pressed, a figure created based on the knowledge read from the storage device 31 in which events are arranged according to the presence or absence of a relationship is displayed. , C3A in the central part 110 of the screen. In addition, if only a part of the knowledge is displayed on the screen, it is possible to scroll one screen.The knowledge editor corrects the knowledge based on the display.6For example, the event type of event O1 in the diagram is
To correct S (State) J, first move the mouse cursor to Off and press the button, then move the cursor to 2 icon 81 and press the button. According to this instruction, C
8A Expressed knowledge is modified and C8 on one screen
The diagram in A is modified. In this way, by performing the work of modifying the diagram displayed on the screen, the knowledge is also automatically modified, making it possible to easily perform one editing work.

Deleting and adding events and links can be easily carried out as described above by specifying the position with the mouse cursor and inputting commands with the buttons. Also, supplementary knowledge names, events, etc.

Corrections can be made by specifying a location using the mouse and inputting a character string from the keyboard.

As explained above, if the device according to this embodiment is used,
A plant state that is relatively easily obtained from a plant, etc., and that results from an operation and its implementation.

In other words, based on event-related data, it is possible to create knowledge about event types and relationships between events for use in plant operation consulting systems, etc. Furthermore, knowledge can be efficiently displayed by displaying the knowledge graphically and inputting it using a mouse, etc. can be edited and corrected.

Note that here, we have explained the operation of the device using specific input data and knowledge expressed in C8A format as an example.
Even when general data about events is used as input, or when knowledge is expressed in an if-then rule format, frame format, etc., it can be done in exactly the same way by making slight changes to the event-related rules or the display screen. The operation and effect can be obtained.

FIG. 9 is a block diagram showing another embodiment of the apparatus of the present invention, including a plant dynamic characteristics simulator 120, a simulated operation control panel 130. Data input device 140, input data record bag [150, driving consulting device 160. ! ! , response content recording device 151, calculation bag M21. Storage device 3
2,33. Input device f141. It consists of a display device [51]. In addition to the calculation unit 21a, the calculation device 21 includes a data processing unit 21b that functions as an interface with various external devices, an event-related rule search unit 210, a search result input unit 21d, and an intelligence output unit 21e. .

There is a knowledge input section 21f, an image data output section 21g, an input command processing section 21h, and a recording data input section 21i.

Furthermore, the storage device @32 stores event-related rules that specify relationships between events. On the other hand, storage device 1t33
1 display device! Necessary to create a graphic signal to 51,
Knowledge data regarding interrelationships between events is stored.

Next, the operation of the apparatus according to this embodiment will be explained.

The plant dynamic characteristic simulator 120 is activated by a command from the command input device 1t41. After starting the simulator, an operation expert such as a plant operator is made to use the simulated operation control panel 130 to perform operations according to a predetermined situation such as a regular test or response to an abnormality.

The data input device 1140 receives operation signals from the simulated operation control panel 130, signals regarding system/equipment status from the plant dynamic characteristic simulator, and signals regarding plant state quantities in real time, and from the state quantities, The arithmetic device 21 performs input processing such as determining the change (increase, decrease, etc.) and sends the information to the data processing unit 21b of the arithmetic device 21, similar to the device of the first embodiment shown earlier. The processing is performed in real time, and the created knowledge is stored in the storage device 33.

After completing a series of operations, the arithmetic device! 21 reads the knowledge from the storage device 33 via the knowledge input section 21f, and uses it to display the information on the display device! ! 51, the events are arranged and displayed graphically according to the presence or absence of a relationship, the knowledge created in 1 is edited and corrected by the same method as the device of the first embodiment, and is sent to the storage device via the knowledge output unit 21ψ. 33.

Furthermore, in the device according to this embodiment, the driving consulting device 160 uses the knowledge stored in the storage device 33.
Ask and answer questions about driving methods. This consulting is based on the plant dynamic characteristics simulator 120.
It is carried out in real time with the system operating. And the question/
The contents of the response are recorded in the storage device 1151. on the other hand,
The flow of plant state changes, operations, etc. is recorded by the input data recording device 1150.

After the predetermined operation is completed, the input data recording device 150
, and the recorded contents of the question/answer contents recording device 151 are displayed on the display device through the recorded data input section 21i and the image data output section 21g. ! 51 is displayed.

Use this as a reference to improve your knowledge. Also, at this time, if you specify it by input, it will be based on the new input data. It is possible to create new knowledge, and it is also possible to reflect this in improving knowledge.

According to the device of this embodiment, in addition to creating knowledge,
It also becomes possible to easily improve knowledge through the use of knowledge.

Note that each of the embodiments of the knowledge base creation support device described above is effective not only for the operation of a plant but also for collecting and maintaining knowledge regarding the operation of an aircraft, a ship, a spaceship, and the like. Furthermore, the knowledge created using the device according to the present invention includes various information related to driving, such as operating procedures and effects of operations, and can be used in driving automation devices and the like. This can be realized by sequence control based on this information or knowledge engineering applied control.

〔Effect of the invention〕

As explained above, according to the present invention, it is necessary for a plant operation consultation system and the like. This has the effect of reducing the amount of work required for gathering and maintaining knowledge related to plant operation methods.

In addition, since manual intervention during knowledge creation can be reduced, human errors can be reduced, and therefore, the use of the apparatus of the present invention has great effects in improving economic efficiency and safety.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing a first embodiment of the present invention;
The figure is a flowchart showing an overview of processing in the arithmetic unit, Figure 3 is a diagram showing an example of input data, Figure 4 is a diagram showing an example of the common sense algorithm (CS A), and Figure 5 is a diagram of C8A. 6 is a flowchart showing a specific example of knowledge creation processing, FIG. 7 is a diagram showing an example of event-related rules, FIG. 8 is a schematic diagram showing an example of a display screen, and FIG. 9 is a flowchart showing a specific example of knowledge creation processing. FIG. 2 is a block diagram showing a second embodiment of the present invention. 10...Data input device, 20.21... Arithmetic device. 20a, 21a...=calculation unit, 20b, 21b--data processing unit, 20ce21c...event-related rule search unit. 20 d, 21 d...Search result input section, 20
e, 21e...knowledge output section, 20f, 21f...knowledge input section. 20 g #21 g...Image data output section, 20
h. lLh...Input command processing unit, 21i...Record data input unit, 30-33...Storage device, 40a...Mouse. 40b...Keyboard, 41...Command input device, 5
0゜A51...Display device, 81-83.../Icon, 90
...Mouse cursor, 120...Plant dynamic characteristics simulator, 130...Simulation operation control panel, 140...
・Data input device, 150... input data recording device,
151... Question/answer content recording device, 160... Driving consulting device.

Claims (1)

[Claims] 1. An input device for inputting plant operation items and a plurality of plant states caused by the implementation of the operation items, and a first memory for storing rules indicating relationships between operation items and plant states and relationships between plant states. and related search means for searching the rules in the first memory to determine the relationship between the operation items inputted from the input device and the plurality of plant states, and the operation items and the plurality of plant states obtained by the means. The presence or absence of a relationship between states is stored as data for graphically displaying, or when a correction part for the data is instructed from outside, the result of correcting the data based on the specified correction part and contents is displayed graphically. a second memory for storing data for display; and a calculation means for calculating a graphical signal illustrating the relationship between the input operation items and the plurality of plant states based on the data stored in the second memory. , a knowledge base creation support comprising: a display device for displaying the graphic signal obtained by the calculation means; and a means for instructing the modified portion and the content of the modification of the graphic signal displayed on the display device. Device.