JPH02181298A - Trouble detecting using image type knowledge base - Google Patents

Abstract

PURPOSE:To easily detect a trouble by describing technical knowledge concerning the detection of a plant trouble, etc., on a drawing in the form of a knowledge structure tree to input the knowledge as image data and recording the route of inference as connection information on node coordinate positions. CONSTITUTION:When a plant trouble occurs, an operator searches the item of the trouble contents via a CRT device 1 to select it from a menu by using a mouse 3. In this case, a computer 2 receives an input signal and displays a knowledge structure tree 7 for trouble diagnosis stored previously with an input operation on the screen of the device 1. A circle mark 7f is added to the coordinate position of each node and the color of this mark is changed to show the position of the present inference. The items to be checked next are described in characters into the frame of each of nodes 7a - 7e. Thus the operator inputs the signal into the mark 7f in the frame of each node by using the mouse 3 when the contents of characters are confirmed. These operations are successively repeated to perform the inference for detection of a trouble.

Description

[Detailed description of the invention]

[Industrial Application Field 1] This invention relates to a fault finding device using an image-based knowledge base that performs fault finding processing in plant systems and training of fault finding operators for non-experts. be. [Prior Art] Fig. 3 is a block diagram showing a conventional video-based knowledge-based fault finding device disclosed in, for example, Japanese Unexamined Patent Publication No. 63-761. In the figure, 2 refers to fault finding. 5 is a television screen on which images, text, and drawings are displayed; 6 is a video disk device equipped with a video of the equipment that corresponds to the procedures for finding faults; 8 is a computer equipped with the expert's rules; A touch panel 9 for interacting with this knowledge-based fault finding device is a superimposer for displaying characters on the television screen 5. Next, the operation will be explained. First, when a failure occurs in a plant system or the like, an operator selects an item indicating the nature of the failure from the television screen 5 by touching the appropriate location on the touch panel 8 with his or her finger. At this time, when the computer 2 receives the input signal from the touch panel 8, it gives an output command to the superimposer 9 that displays characters and the video disc device 6 in accordance with a pre-stored program of rules regarding fault diagnosis. Then, the superimposer 9 gives an output signal to the television screen 5 that superimposes the two screens of the image of the video disk device 6 and the instructions for the fault finding operation, and displays them simultaneously. The operator discovers and restores failures in the plant system based on the instructions for the failure finding operation, but during the process, the operator discovers things that are unclear among the things that the operator should check (for example, where to check,
How to find out 1) If there is a word (meaning, etc.), touch panel 8
By touching the corresponding location above with your finger, a video of the answer (guidance) will be displayed on the television screen 5. In addition, video images that should be the answer (for example, the entire equipment, the exterior of the panel, the interior of the panel, relays, chassis cards, parts, etc.)
All the images necessary for the process of fault discovery by experts are stored in the video disc device 6 in advance.

[Problem to be solved by the invention]

Conventional video-based knowledge-based fault finding equipment is configured as described above, so in order to input expert knowledge into this knowledge-based fault finding equipment, the contents of that knowledge are input into a program as a knowledge base. This is necessary, but it requires a lot of time. Moreover, there is a problem in that it is not easy to modify or add rules for performing knowledge-based reasoning. This invention was made in order to solve the above-mentioned problems, and when constructing a knowledge base, the input work of inputting a huge amount of expert knowledge into a fault detection device by an operator can be done without knowledge of programming. The purpose of the present invention is to obtain a failure detection device based on an image-based knowledge base that can be easily performed and rules can be easily modified and added.

[Means to solve the problem]

The fault finding device using an image-type knowledge base according to the present invention organizes the knowledge of experts as a tree-like knowledge structure tree and inputs what is written on a drawing into a computer via an image input device as image data. , the connection status between each node of the knowledge structure tree displayed on the display device is input via the coordinate specifying device. Further, during a fault finding operation, instructions regarding the fault finding operation indicated by the knowledge structure tree and its connection status are displayed on the display device to instruct fault finding.

[For use]

Inputting expert knowledge to the fault finding device in this invention is performed by inputting a drawing in which knowledge organized as a tree-like knowledge structure tree to a computer using an image input device. In addition, by selecting and instructing the relationship between the node parts of the tree-like knowledge from the display device via the coordinate specifying device, in other words, by associating the coordinate positions on the image data, inference for failure discovery is advanced, and if necessary, the Find the malfunction by referring to the screen and operating sequentially in an interactive manner.

【Example】

An embodiment of the present invention will be described below with reference to the drawings. In FIG. 1, the same parts as in FIG. 3 are designated by the same reference numerals. In FIG. 3 is a mouse (coordinate designation device) which is a pointing device; 4 is an image input device for reading and inputting image data such as drawings;
For example, an image scanner. Next, the operation will be explained. First, in the same way as before, the video disk device 6 is equipped with video and specialized information that will be needed in the process of fault detection of the entire equipment, the exterior of the panel, the inside of the panel, relays, chassis cards, parts, etc. A collection of video images showing the methods of investigation carried out by houses has been accumulated. Next, inputting specialized knowledge related to failure detection will be explained. The specialized knowledge related to failure discovery is organized into a tree, and this is called a knowledge structure tree 7. The knowledge structure tree 7 is traced on the drawing in the form shown in FIG. In the example in Figure 2, 7a is the starting point of the inference, 7b is the branch (necessary question for the inference), 7C is the instruction for investigation, and 7d is the conclusion, and the necessary information is shown in text within each frame. has been done. Since the knowledge structure tree 7 needs to be displayed on the screen of the CRT device 1, if it becomes larger than the screen, it will be divided into multiple pieces, and in FIG.
It's raining in Shigemaru. A drawing describing this knowledge structure tree 7 is directly input as image data through the image input device 4 shown in FIG. 1, and is stored in the computer 2. The knowledge input person is
The image data of the knowledge structure tree 7 is displayed on the CRT device 1, and each node 7a is displayed according to the message from the system.
, 7b, 7c. Enter the connection status between 7d and 7C using mouse 3. The computer 2 selects each node 7a7b of the knowledge structure tree 7,
The connection status between 7c, 7d, and 7C is stored as the connection status between the coordinate position data specified by the mouse 3. Note that the connection information between each node here indicates a route for finding a fault. Next, the actual operation for finding a fault is performed as follows. In other words, when a failure occurs in the plant, the operator
From the RT device 1, find the item indicating the failure details and select it from the menu using the mouse 3. At this time, the computer 2 receives this input signal and uses the fault diagnosis knowledge structure tree 7 that has been accumulated in advance through the previous input operation.
is displayed on the screen of the CRT device 1. At this time, the coordinate position of each node recorded by the previous input operation is
A circle mark 7r is displayed, and the current inference position is indicated by the color change of this circle mark. The next thing to check is each node 7a,
The characters are written in the frames 7b, 7c, 7d, and 7e, so when the operator confirms the content of the characters, he uses the mouse 3 to send a signal to the circle 7f in the frame of each node. input. By sequentially repeating the above operations, inference for fault discovery is performed. In addition, during inference, if there is something that the operator is unsure about (e.g., where to check, how to check, the meaning of the five words, etc.), the operator can use the mouse 3 to point to the menu on the CRT device 1. If the answer is a video image, it will be displayed on the TV screen 5, or if it is a drawing (image data), it will be displayed on the C
displayed on the RT device 1. Note that in the above embodiment, a circle 7[ is used as a symbol indicating the position of a node, but it may be a square or a triangle, and is not particularly limited. Further, there is no particular restriction on the marking of the frame depending on the meaning of the node, and any shape may be used. Also. In the above embodiment, a procedure for finding a failure in a plant or the like has been described, but the same effects as in the above embodiment can be obtained even if the present invention is used for process diagnosis, control guidance, etc. of various plants. [Effects of the Invention] As described above, according to the present invention, specialized knowledge related to plant failure detection, etc. is described as a knowledge structure tree on a drawing, this knowledge structure tree is input as image data, and , Since the path of inference is recorded as connection information of coordinate position data of nodes on image data, large amounts of specialized knowledge data can be input easily and quickly without any knowledge of programming, and it is also possible to modify knowledge. This has the effect of making additions easy.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing a knowledge-based fault finding device according to an embodiment of the present invention, FIG. 2 is a model diagram of a knowledge structure tree according to the present invention, and FIG. 3 is a diagram showing a conventional knowledge-based fault finding device. FIG. 1 is a CRT device (display device), 2 is a computer, 3 is a mouse (coordinate designation device), 4 is an image writing device, 5
is a television screen, 6 is a video disc device, and 7 is a knowledge structure tree. In addition, in the figures, the same reference numerals indicate the same or equivalent parts. (2 other people) Figure 7: Ooguchi 8-on and Doubushigeishi

Claims (1)

[Claims] Commands from a computer equipped with expert knowledge about fault finding are transmitted to a video disk device that has images of the equipment where the fault was discovered, which are associated with the fault finding procedure, and the images, text, drawings, etc. are transmitted to a television. In a fault finding device based on an image-based knowledge base that is displayed on a screen and is displayed on a screen to discover faults in the device in an interactive manner with an operator, the system is connected to the computer and relates to image data of a knowledge structure organized in a tree shape and the operation of the fault finding device. a display device for displaying instructions; a coordinate specifying device connected to the computer for specifying coordinates within a display screen of the display device; and an image data input device connected to the computer for reading and inputting the image data. A fault finding device based on an image-based knowledge base.