JP2697160B2 - Knowledge base processor - Google Patents

Description

Description: Object of the Invention [Industrial application field] The present invention relates to a knowledge base processing apparatus employed in a CAM system that determines processing know-how from a raw material to a product based on a knowledge base. .

4. [Conventional Technology] In recent years, research and development of knowledge-based systems for solving problems by effectively using specialized knowledge have been actively performed, and such knowledge-based systems have been employed in CAM systems.

An example of such a knowledge base system is shown in FIG. As shown in the figure, this system includes a knowledge base A1 in which machining know-how is accumulated, and a control unit A2 that determines optimal machining know-how from the knowledge base A1. Knowledge Base A1 states, "The findings of (for example, the material is iron)
If it is, it is a set of rules such as "The processing procedure of ~". In detail, the finding factor, that is, the decision factor which sets H7 Toshi (hole drilling processing information) or iron, and the decision factor There are provided a plurality of tables A11 composed of processing know-how determined based on the information. The control unit A2 selects a desired table from the knowledge base A1 (A2
1) Extract the table A11 from the knowledge base A1 (A2
2) Then, from the CAM database A3 in which various data required for the processing of the coarse material together with data indicating the processed product shape are stored, all the factor information used for the judgment of the finding is extracted (A23). A22
(A24), and when all the factor information coincides with the above finding, the optimal processing know-how is determined (A25).

[Problems to be Solved by the Invention] By the way, the construction of such a knowledge base system involves:
It is very important how to apply the knowledge of experts, but because of the difficulty, it is necessary to add knowledge base data,
Maintenance to correct and erase is inevitable. However, in the above-described conventional technique, it is necessary to frequently change the size of the table in order to perform maintenance,
There was a problem that the man-hour and period required for the maintenance were enormous.

Furthermore, since maintenance is difficult as described above, it is necessary to thoroughly investigate the knowledge of experts before developing a knowledge-based system, and the problem that construction of a knowledge-based system takes a long time is also invited. It had been.

The present invention has been made in view of these problems,
An object of the present invention is to provide an excellent knowledge base processing apparatus that can easily maintain a knowledge base in which processing know-how is collected in a short period of time.

Configuration of the Invention [Means for Solving the Problems] In order to achieve the above object, the present invention has the following structures as means for solving the above problems. That is, as shown in FIG. 1, a knowledge base processing device of the present invention is provided in advance with a knowledge base M1 which collects rules for determining processing know-how from a crude material to a product from given findings. A control unit M2 for finding the finding based on the product information to which the rough-shaped material information has been given, and determining optimal processing know-how from the knowledge base M1. One rule of M1 is a plurality of decision factors M1 including a header part indicating the judgment element of the finding and a data part indicating the judgment result of the judgment element.
And one conclusion factor M12 indicating machining know-how corresponding to the contents of the plurality of decision factors M11, and the control unit M2 includes a rule extracting unit M21 for extracting one rule from the knowledge base M1. A determination factor extracting means M22 for sequentially extracting the determination factor M11 of the extracted one rule; and a determination element indicated in a header portion of the extracted determination factor M11 in the product information to which the coarse material information is added. Determining means M23 for determining whether or not the corresponding partial information matches the determination result indicated in the data part of the decision factor M11; and all decision factors M11 included in one rule extracted by the rule extracting means M21. When it is determined that the determination results of the determination means M23 match over a period of time, the processing know-how indicated by the conclusion factor of the one rule is maximized. It is characterized in that it comprises a determining means M24 for determining the Do processing know-how.

[Operation] In the knowledge base processing device of the present invention configured as described above, one rule of the knowledge base M1 is composed of a plurality of headers each indicating a judgment element of a finding and a data section indicating a judgment result of the judgment element. Of the decision factor M11, and one conclusion factor M12 indicating machining know-how corresponding to the contents of the plurality of decision factors M11.The control unit M2 determines the optimal machining know-how from the knowledge base M1. However, in detail, one rule is extracted from the knowledge base M1 by the rule extracting means M21, and the extracted one rule is extracted.
The decision factor M11 possessed by the rule is sequentially taken out by the decision factor taking out means M22, and the partial information corresponding to the judgment element shown in the header part of the taken out decision factor M11 in the product information to which the coarse material information is added is The determination means M23 determines whether or not the determination result M11 matches the determination result indicated in the data part of the extracted decision factor M11. When it is determined that the determination results of the determination unit M23 match, the processing know-how determining unit M24 determines the processing know-how indicated by the conclusion factor M12 of the one rule as the optimum processing know-how.

Therefore, since the knowledge base is constructed using the decision elements of one rule and the decision results as independent decision factors, maintenance of the knowledge base M1 such as addition or deletion of decision elements in one rule is performed. Also, maintenance can be performed without changing the entire data structure by simply adding or deleting decision factors.

Next, a preferred embodiment of the present invention will be described in detail.

FIG. 2 is a block diagram showing a schematic configuration of a CAM system for machining equipped with a knowledge base processing apparatus according to one embodiment of the present invention.

As shown in the figure, this machining CAM system
It is composed of a general-purpose computer and its peripheral devices, and has a keyboard 1 for inputting characters and various instructions by keys, a graphic input device (tablet) 2 for inputting graphics, and a display device (CRT display) 3 for displaying characters and graphics. A tape puncher 4 for punching calculation results on a paper tape, disk devices 5 and 6 as external storage devices for storing various data, and an electronic control device 10 connected to these devices for performing various arithmetic processing. I have.

The disk device 5 stores a product information model MDL, which is designed and input by CAD (not shown) and indicates a processed product shape.

In the disk device 6, a knowledge base (knowledge base) KB, which is a set of rules, is stored. The knowledge base KB has a machining procedure base KB1 that is a set of rules for determining a machining procedure, a tool selection base KB2 that is a set of rules for selecting a tool, and a tool selection base KB2 that has a tool path. Tool path determination base which is a set of rules
It consists of KB3.

The knowledge base KB is stored in the disk device 6
More specifically, as shown in FIG. 3, the operator key-in various rules prepared in advance from the keyboard 1 to generate a machining procedure base KB1, a tool selection base KB2, and a tool path determination base KB3. The data is classified and stored in the disk device 6.

The data structures of the machining procedure base KB1, the tool selection base KB2, and the tool path determination base KB3 will be described below.

As shown in FIG. 4, the processing procedure base KB1 stores one rule in a separate line, and the one rule is
Multiple decision factors consisting of header part HD and data part DT
FA and one conclusion factor FA0 indicating machining know-how corresponding to the contents of the plurality of decision factors FA. The header part HD of the deciding factor FA stores data indicating the judgment factor of the finding of knowledge, and the data part DT of the deciding factor FA stores data indicating the judgment result of the judging element. I have. Here, the judging element is a machining shape name material and the like, and the judgment result is that the machining shape name is H7 toshi or the material is iron. Then the conclusion factor FA0
Defines the machining know-how corresponding to the content of these decision factors FA. The machining know-how includes a machining procedure for shaving a product shape from a rough material, for example, FM (face milling) -CE (centering). ) -DR (drilling) -RM (reaming) -TP (tapping). Incidentally, the conclusion factor FA0 is actually divided into a header part and a data part like the decision factor FA, and the type of the processing know-how, for example, the processing procedure is stored in the header part, and the data part is stored in the data part. A conclusion for the type, for example, FA-CE-DR-RM-TP is stored. In addition, the conclusion factor FA0 indicates a continuation or a termination.] (*: Continuation, ◎: termination) is written, and if the continuation * is written, the next line Is further inferred with the contents of the determination factor FA, and the processing know-how is determined by the conclusion factor FA0 having the terminator ◎. FIG. 4 shows a storage example in which: if A = a1 and B = b1, Y = y1; if Y = y1, C = c1, and D = d1, X = x1.

Tool selection base KB2 and tool path determination base K
Similarly to the processing procedure base KB1, B3 also includes a plurality of decision factors FA including a header part HD and a data part DT, and one conclusion factor FA0 indicating processing know-how corresponding to the contents of the plurality of decision factors FA. It is configured.

On the other hand, the electronic control unit 10 includes a well-known CPU 11, a ROM 12, and a RAM 1.
It is configured as an arithmetic logic circuit centering on 3 and the like, and includes an input / output circuit 14 for inputting / outputting data from / to the peripheral device described above, a communication port 15 for receiving data sent from the outside, and the like. By executing various arithmetic processes in the electronic control device 10, NC data is created, and the NC data is supplied to the outside via the tape puncher 4 in the form of an NC tape.

Hereinafter, the electronic control unit 10
The various calculation processes executed by the above will be described.

As shown in FIG. 3, when the knowledge base is stored in the disk device 6 in the preparation work, the actual work is performed thereafter.

As actual work, first, input processing of product information and coarse material information is performed. More specifically, a product information model MDL designed by an external CAD is taken in via the communication port 15 and stored in the disk device 5 to define the product (FIG. 3). The rough material information is defined by adding the rough material information input through the disk device 2 to the product information model MDL stored in the disk device 5 (FIG. 3). In addition,
The rough material information is information necessary for processing the rough material, and corresponds to, for example, a material, presence or absence of casting, rigidity, and the like.

Subsequently, the process proceeds to a process design process. First, the operator receives a designation via the keyboard 1 and the tablet 2 on which processing machine (machine) the processing is to be performed (FIG. 1). Thereafter, an instruction is given from the operator as to which machining part the machining procedure and the tool are automatically determined (FIG. 1). This instruction is issued for all the processing parts having the same shape, only by making one instruction for the processing parts having the same shape. Subsequently, referring to the machining procedure base and the tool selection base stored in the disk device 6, an optimal machining procedure and a tool are automatically determined for each machining portion, and the determination result is stored in the disk device 5. Perform processing for automatic tool determination calculation (FIG. 2).

Subsequently, the process proceeds to the processing of tool use order / tool path determination. First, an operation of automatically determining the use order of the tool for processing each part is performed (FIG. 3). More specifically, the order in which the tools used for each part, which have been automatically determined and calculated, are used most efficiently is automatically determined, and the determination result is stored in the disk device 5. Thereafter, an operation of determining a movement path (cutter path) of the tool for processing each of the parts is performed (FIG. 6). In detail, based on the product information model MDL stored in the disk device 5 and to which the coarse material information is added, the moving path of the tool is determined with reference to the machining procedure base KB1 stored in the disk device 6, and the determination is made. The result is stored in the disk device 5.

Thereafter, the operation of automatically creating NC data based on the various data stored in the disk device 5 is performed as described above, and the NC data is output to the outside from the tape puncher 4 using a paper tape as a medium (FIG. ).

Next, how to apply the knowledge base KB will be described in detail with reference to an example of a routine for calculating a processing procedure in the above processing. In this routine, the processing is performed on the processing part specified in the above processing.

As shown in the flowchart of FIG. 5, when the processing is started, first, from the processing procedure base KB of the knowledge base KB to 1
The rule is taken out (step 100). Next, one factor is extracted from the extracted one rule (step
110) Then, it is determined whether the extracted factor is a decision factor FA or a conclusion factor FA0 (step 120).

If it is determined in step 120 that the factor is the determination factor FA, it is determined whether or not data relating to the determination factor indicated in the header HD of the determination factor FA is in a data pool area described later (step 130). Here, when it is determined that there is no data, the data relating to the determination element is extracted from the designated predetermined processing portion of the product information model MDL stored in the disk device 5 and to which the coarse material information is added (step S1). 140), and the extracted data is
It is determined whether or not it is equal to the determination result indicated in the data part DT of the decision factor FA extracted in (step 150).
On the other hand, if it is determined in step 130 that the data related to the determination element is in the data pool area, the processing in step 140 is skipped, and the process proceeds to step 150 where the data related to the determination element in the data pool area is skipped. Is the data part D of the decision factor FA extracted in step 110
It is determined whether it is equal to the determination result shown in T.

If it is determined in step 150 that they are equal, the process returns to step 110, extracts the subsequent factor from the one rule extracted in step 100, and repeats the subsequent processes. On the other hand, if it is determined in step 150 that they are not equal, the process returns to step 100, fetches a new rule, and repeats the subsequent processes.

If it is determined in step 120 that the extracted factor is the conclusion factor FA0, the process proceeds to step 160.
Proceed to. In step 160, it is determined whether or not the conclusion factor FA0 has a terminator.
That is, when it is determined that the rule has a continuator, the contents of one rule extracted in step 110 are stored in the data pool area in the RAM 13 (step 170).
And a new rule is taken out and the subsequent processing is repeated. The contents of one rule stored in the data pool area in this way are the same as those described above when the AND condition is added to the judgment element in the previous one rule in the new one rule that has been taken out. 150 will be used. On the other hand, if it is determined in step 160 that it has a terminator, then the determination factor indicated in the header HD of the conclusion factor FA0 and the determination result indicated in the data part DT are calculated as a calculation result and stored in a predetermined area in the RAM 13. It is temporarily stored (step 180). After the execution of step 180, the processing of this routine ends once.

Therefore, according to the processing procedure automatic calculation routine configured as described above, the rule extracted in step 100 is, for example, as shown in FIG.
If the material is iron and the finishing procedure is DR-BR (boring) -RM if the material is iron, step 110
First, the determination factor FA that the machining shape is H7 toshi is extracted, and it is determined that it is the determination factor FA in step 120. Since nothing is initially stored in the data pool area, a negative determination is made in step 130. And
In step 140, the processing shape at the designated predetermined processing portion of the product information model MDL is extracted, and the determination result stored in the data part DT of the determination factor FA, that is, whether the extracted processing shape is equal to H7 toshi is determined. When the processed shape is H7 toshi as shown in FIG. 6 (b), it is determined that both are equal. Subsequently, in step 110, a determination factor FA such as that the material is iron
Is extracted, and the process proceeds to steps 120, 130, 140, and 150. In step 150, an affirmative determination is made again, and as a result,
Returning to step 110 again, a conclusion factor FA0 such that the machining procedure is DR-BR-RM is extracted, and as a result, a determination is made at step 120 as a conclusion factor FA0, and the process proceeds to step 160, where the conclusion factor FA0 is determined. Has a terminator ◎, so a positive determination is made in step 160,
In step 180, the contents of the conclusion factor, that is, the processing procedure is DR-BR-RM, is stored in the data RAM13.

According to the knowledge base processing apparatus of the present embodiment described in detail above, the knowledge base KB is constructed using the decision element of one rule and the decision result as an independent decision factor FA. Addition of the judgment element in,
Maintenance such as erasure can be performed simply by adding or erasing the decision factor FA without changing the entire data structure.
It can be done easily. Therefore, the man-hour and period required for the maintenance can be greatly reduced. In addition, since maintenance is easy in this way, it is possible to start system development without investigating all the judgment factors of knowledge, and as a result, the period until system startup can be greatly reduced.

In addition, since such a CAM system does not require an expensive dedicated computer for Al or a dedicated language such as Prolog or Lisp, a general-purpose computer can be constructed in a language,
The time until the development and practical application of the knowledge base can be dramatically reduced.

As mentioned above, although one Example of this invention was described in full detail, this invention is not limited to the said Example at all, and can be implemented in various aspects in the range which does not deviate from the summary of this invention. Of course it is.

Effects of the Invention As described in detail above, according to the knowledge base processing apparatus of the present invention, it is possible to easily perform maintenance of a knowledge base in which processing know-how is collected, and to significantly reduce the number of man-hours and period required for maintenance. it can. In addition, since maintenance is easy in this way, it is possible to start system development without investigating all the judgment factors of knowledge, and as a result, the period until system startup can be greatly reduced.

[Brief description of the drawings]

FIG. 1 is a basic configuration diagram illustrating a basic configuration of the present invention,
FIG. 2 is a block diagram showing a schematic configuration of a CAM system for machining equipped with a knowledge base processing apparatus according to an embodiment of the present invention. FIG. 3 is an explanatory diagram showing operations performed by the CAM system. FIG. 4 is an explanatory diagram showing a data structure based on a machining procedure which is one of the knowledge databases, FIG. 5 is a flowchart showing a machining procedure automatic calculation routine executed by the electronic control unit, and FIG. FIG. 6B is an explanatory diagram showing an example of a procedure-based data content, FIG. 6B is an explanatory diagram showing an example of a product information model, and FIG. 7 is an explanatory diagram showing a conventional example. M1 ... Knowledge base, M11 ... Determining factor M12 ... Conclusion factor, M2 ... Control unit M21 ... Rule extracting means M22 ... Determining factor extracting means M23 ... Determining means, M24 ... Determining means 1 ... Keyboard …… Tablet 3… CRT display 5,6… Disk device, 10… Electronic control device MDL …… Product information model KB …… Knowledge base, KB1 …… Machining procedure base KB2 …… Tool selection base KB3… … Tool path determination base

 ────────────────────────────────────────────────── ─── Continuation of front page (72) Inventor Kuniharu Muramatsu 1 Toyota Town, Toyota City, Aichi Prefecture Inside Toyota Motor Corporation (56) References JP-A-62-226267 (JP, A)

Claims (1)

(57) [Claims] 1. A knowledge base which collects rules for determining processing know-how from a crude material to a product based on a given finding, and a product information to which coarse material information prepared in advance is added. A control unit that finds a finding and determines optimum machining know-how from the knowledge base, wherein one rule of the knowledge base is a header part indicating a judgment element of the finding and the judgment. A plurality of decision factors each including a data part indicating a result of determination of an element, and one conclusion factor indicating processing know-how corresponding to the contents of the plurality of decision factors. Rule extracting means for extracting one rule from the rule, determining factor extracting means for sequentially extracting the determining factor of the extracted one rule, and the coarse material information It is determined whether or not the partial information corresponding to the determination element shown in the header part of the extracted decision factor in the product information to which the determination factor is given matches the determination result shown in the data part of the decision factor. A determining unit that, when it is determined that the determination results of the determining unit match over all the determining factors of the one rule extracted by the rule extracting unit, the processing indicated by the conclusion factor of the one rule A determination means for determining know-how as optimal processing know-how.