JP2006236299A - Integrated knowledge based system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an integrated knowledge information system capable of copying with a situation request in which information over many areas increases explosively and which simultaneously requires an urgent solution about a disaster, an accident or the like. <P>SOLUTION: In this knowledge based system comprising a host computer and an inputting/outputting part and/or a display part connected to the host computer, the inputting/outputting part displays (outputs as a character, an image, a voice or printed information) an answer to an inquiry on the basis of information on a presentation layer, uses knowledge obtained by analyzing the following knowledge such as data, a diagram, a drawing, a table, a relation, a procedure, a rule, a calculation module and knowledge included in various technical standards or the like, represents the knowledge with a general-purpose information model being a unified common expression, semiautomatically converts the knowledge into formal knowledge, stores an integrated knowledge base, reuses and analyzes the integrated knowledge base, and provides knowledge information necessary to the entire life cycle such as the design of a plant, etc., handling and operation, preventive maintenance, disaster countermeasures or the like, providing the environment of a rational and efficient knowledge use. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

The present invention relates to a knowledge base system. More specifically, the present invention is a system for gathering technologies necessary for constructing a knowledge base, and various forms of information, that is, character data, tables, diagrams, mathematical formulas, procedures, etc. The present invention relates to a knowledge base system or a knowledge base system that collects the knowledge contained in the database and enables reuse.

Today, with the development of industry and science, there is a need to solve problems by integrating various forms of knowledge, such as drawings, across multiple fields. To this end, useful information is accurately identified and collected from large-scale and distributed information sources, useful knowledge is efficiently acquired from a variety of information sources, and feedback from users and changes in information are quickly and accurately determined. Knowledge-based technology that is reflected in knowledge is important.

For example, to ensure the safety of chemicals and power plants, even if it is a part of work like pipe management, fluid dynamics, thermodynamics, chemical engineering, materials, machinery, corrosion, wear, thermal stress, piping, control A large amount of specialized information and knowledge over a wide area such as architecture and iron structures are related.
Furthermore, information and knowledge necessary for solving these problems are in various forms from various information sources, and useful knowledge must be acquired efficiently from now on.

Furthermore, information and knowledge necessary for solving these problems are in various forms from various information sources, and useful knowledge must be acquired efficiently from now on. For example, technical information and knowledge related to equipment are included in design drawings such as flow sheets and in-house documents such as equipment specifications. Information and knowledge that can be used more effectively and usefully may be included in domestic and international industrial standards such as JIS and ANSI, and standards such as design standards and drafting standards within a company. These technical documents include, in addition to the so-called knowledge as rules described in text documents, the designer's intentions included in the drawings, knowledge on constraints such as allowable pressures expressed in tabular form, etc. .

However, data, mathematical formulas, and procedures are treated as out-of-frame in today's artificial intelligence and knowledge engineering frameworks. Moreover, even though the diagrams and tables contain a high level of knowledge, such as the designer's intentions, only humans can understand them, and they were not used directly in computer systems. It is necessary to clarify each and share it as formal knowledge that can be used by both humans and computers and reuse it.

However, data, mathematical formulas, and procedures are treated as out-of-frame in today's artificial intelligence and knowledge engineering framework. Moreover, even though the diagrams and tables contain a high level of knowledge, such as the designer's intentions, only humans can understand them, and they were not used directly in computer systems. For this reason, although it is necessary to use various forms of knowledge in various fields to solve problems, there are many systems in which individual knowledge is implemented. There wasn't.

JP-A-4-326167 JP-A-4-326428 JP-A-6-83678

Academic literature 1

General-purpose model (GPM) and knowledge representation

The present invention has been made in view of the above circumstances, and unlike the prior art, the object of the present invention is to integrate data, diagrams, tables, mathematical formulas, procedures by converting them into common formal knowledge and storing them. To build and provide a knowledge base system.

Another object of the present invention is to integrate all the knowledge contained in rules, data, diagrams, tables, mathematical formulas, procedures, etc. by using a common format, thereby enabling human beings and companies that are groups of them. It is also possible to support intellectual activities aimed at sustaining and maintaining the existence of society and society through the network.

The object of the present invention is to convert the rules, relations, data, and procedures into a general-purpose information model that is a common expression, and the knowledge contained in drawings, tables, etc. is the knowledge used when humans understand these. It is intended to provide an environment that makes it easy to reuse knowledge by automatically converting it to formal knowledge using knowledge obtained by analyzing the knowledge.

In addition, an object of the present invention is to provide an environment that can logically search for a piping potential risk location in, for example, a petrochemical or power plant and use it for preventive maintenance of the plant in order to prevent similar disasters. is there.

The above object of the present invention is achieved by the following means. However, the description of the computer system for realizing the knowledge base system according to the present invention means its logical configuration, and the physical configuration is not particularly limited. Basically, an operating system or a virtual system that understands a so-called object-oriented language is assumed.
(1) In a knowledge base system comprising a host computer and an input / output unit and / or a display unit connected to the host computer,
The host computer has a class library layer, a knowledge base layer, a knowledge processing layer, and a presentation layer,
The class library layer includes a class library of objects (knowledge objects) in a knowledge representation format expressed by behavior and properties (elements / attributes).
The knowledge base layer has a set of knowledge objects that are used or newly defined and represented by each element knowledge of knowledge expression format objects (knowledge objects) of the class library layer,
The knowledge processing layer has a protocol (procedure) for processing information based on information input from the input / output unit and knowledge in the knowledge base layer or the class library layer,
The presentation layer obtains data necessary for display on the display unit from the knowledge processing layer, and passes information to the input / output unit.
The input / output unit displays an answer to an inquiry based on information in the presentation layer (outputs as text, image, sound, or print information), and is an integrated knowledge base system.
(2) The class library layer or the knowledge base layer is connected to an interface unit with an external database,
The interface unit includes conversion means for converting information from the external database into an object (knowledge object) in a knowledge representation format,
The integrated knowledge base system according to (1), further comprising conversion information storage means for storing the conversion result of the information (object conversion information in a knowledge representation format) in the class library layer or the knowledge base layer.
(3) The knowledge base layer converts analog data information / digital data information such as drawings, diagrams, tables, text data, audio data, or other analog data information / digital data information into knowledge representation objects (knowledge objects). Conversion means for converting to
(1) or (2), characterized by comprising conversion information storage means for storing the conversion result of each piece of information (that is, the result of conversion into an object (knowledge object) in a knowledge representation format) in the knowledge base layer. Integrated knowledge base system described in
(4) In a knowledge base system comprising a host computer and an input / output unit and / or a display unit connected to the host computer,
The host computer has a class library layer, a knowledge base layer, a knowledge processing layer, and a presentation layer,
The class library layer has a class library of objects (knowledge objects) expressed in behavior and properties (elements / attributes).
The knowledge base layer has a set of knowledge objects that use or newly define and represent each element knowledge that is an object (knowledge object) in the knowledge representation format of the class library layer,
The knowledge processing layer has a protocol (procedure) for processing information based on information input from the input / output unit and knowledge in the knowledge base layer or the class library layer,
The presentation layer obtains data necessary for display on the display unit from the knowledge processing layer, and passes information to the input / output unit.
The input / output unit displays an answer to the inquiry based on the information of the presentation layer (output as text, image, audio or print information),
The integrated knowledge base system, wherein the knowledge processing layer includes inference means configured by a procedural language and a logical language.
(5) The knowledge processing layer includes an inference means having a function for calling an external calculation function and using it for a logical operation or a procedural operation, (1), (2), (3 ) Or the integrated knowledge base system according to (4).
(6) In the above (1), (2), (3), (4) or (5), the knowledge processing layer and / or the presentation layer has an interface unit with an external business system. Integrated knowledge base system described.
(7) In a knowledge base system comprising a host computer and an input / output unit and / or a display unit connected to the host computer,
The host computer has a class library layer, a knowledge base layer, a knowledge processing layer, and a presentation layer,
The class library layer includes a class library of objects (knowledge objects) in a knowledge representation format expressed by behavior and properties (elements / attributes).
The knowledge base layer has a set of knowledge objects that use or newly define and represent each element knowledge that is an object (knowledge object) in the knowledge representation format of the class library layer,
The knowledge processing layer has a protocol (procedure) for processing information based on information input from the input / output unit and knowledge in the knowledge base layer or the class library layer,
The presentation layer obtains data necessary for display on the display unit from the knowledge processing layer, and passes information to the input / output unit.
The input / output unit displays an answer to the inquiry based on the information of the presentation layer (output as text, image, audio, or print information), and the interface unit connected to the class library layer or the knowledge base layer includes The second knowledge object conversion means for converting the information from the external database into the XML format by the first knowledge object conversion means for converting the information into the knowledge expression format object and converting it into the knowledge expression object (knowledge object). An integrated knowledge base system characterized by comprising:
(8) It has a conversion means for converting information into an object (knowledge object) in the form of knowledge expression,
Electronic drawings, diagrams, tables, symbols, legends, formulas, functions, procedures, rules, design standards, standards, or industry standards diagrams, tables, symbols, legends, formulas, functions, procedures, rules, standards, etc. Data input / output means;
A user interface that interactively performs a function of converting the data into an object (knowledge object) in a knowledge representation format;
The integrated knowledge base system according to (1), (2), (3), (4), (5), (6) or (7), characterized by comprising:
(9) Electronic drawings, diagrams, tables, symbols, legends, formulas, functions, procedures, rules, design standards, standards, or industry standards drawings, diagrams, tables, symbols, legends, formulas, functions, procedures Input / output means for computerized data such as rules, standards, etc .;
Symbol recognition means for recognizing the digitized data as a symbol;
Display means for displaying the digitized data;
A user interface that interactively performs a function of converting the digitized data into a knowledge representation object (knowledge object);
The integrated knowledge base system according to (7) or (8), characterized in that
(10) Electronic drawings, diagrams, tables, symbols, legends, formulas, functions, procedures, rules, drawings, tables, symbols, legends, formulas, functions, procedures, etc. Input / output means for computerized data such as rules, standards, etc .;
Symbol recognition means for recognizing the digitized data as a symbol;
Topology relation recognition means for recognizing a relation between a plurality of recognized symbols (joint relation, connection relation, connection relation, connection position relation, etc. as a topology relation;
Means for modifying said symbol recognition means and topology recognition means;
Means for converting recognized symbols and topology elements / attributes into knowledge representation objects (knowledge objects);
The integrated knowledge base system according to the above (7), (8), (9).
(11) Electronic drawings, diagrams, tables, symbols, legends, mathematical formulas, functions, procedures, rules, design standards, standards, or industry standards drawings, diagrams, tables, symbols, legends, mathematical formulas, functions, procedures Input / output means for computerized data such as rules, standards, etc .;
Symbol recognition means for recognizing the digitized data as a symbol;
Topology relation recognition means for recognizing a relation between a plurality of recognized symbols (joint relation, connection relation, connection relation, etc. or connection position relation as a topology relation;
Means for modifying said symbol recognition means and topology recognition means;
Means for converting recognized symbols and topology elements / attributes into knowledge representation objects (knowledge objects);
The integrated knowledge base system according to (7), (8), (9), (10), wherein general drawing information is converted into a knowledge expression format object (knowledge object).
(12) In a knowledge base system comprising a host computer and an input / output unit and / or a display unit connected to the host computer,
The host computer has a class library layer, a knowledge base layer, a knowledge processing layer, and a presentation layer,
The class library layer includes a class library of objects (knowledge objects) in a knowledge representation format expressed by behavior and properties (elements / attributes).
The knowledge base layer has a set of knowledge objects that are used or defined and expressed by each element knowledge of knowledge expression format objects (knowledge objects) of the class library layer,
The knowledge processing layer has a protocol (procedure) for processing information based on information input from the input / output unit and knowledge in the knowledge base layer or the class library layer,
The presentation layer obtains data necessary for display on the display unit from the knowledge processing layer, and passes information to the input / output unit.
The input / output unit displays an answer to an inquiry based on information in the presentation layer (outputs as an image, audio, or print information), and is an interface unit connected to the class library layer or the knowledge base layer ? Has a second knowledge objectification conversion means for converting the information from the external database into the XML format by the first knowledge objectification conversion means for converting the information from the external database into an object (knowledge object) in the knowledge expression format. An integrated knowledge base system
In a network environment, the component information in which the host computer, the computer design system, the drawing understanding system, the external calculation system, and the external database are distributed and transferred is transferred in a unified knowledge expression format, and the knowledge base Integrated knowledge base system that accumulates various kinds of knowledge and can be used by remote users.

As described above, according to the present invention, the following effects can be obtained.
In other words, all knowledge contained in rules, data, diagrams, tables, formulas, procedures, etc. can be converted into logical expressions using first-order predicate logic, and expressed and described by knowledge objects in a common knowledge representation format. This enables them to be accumulated as a knowledge base, and by using the knowledge calling procedure prepared by the present invention, effective knowledge contained in a large amount of accumulated data and other information can be obtained. It can be reused.
In addition, as in the case of learning and training program development, we analyze what the elemental abilities of drawing comprehension are based on and analyze the structure between the elemental abilities to build those abilities into the system. By doing so, it becomes possible to effectively use the knowledge limited to the skilled person and the expert so that the same judgment and drawings as the expert can understand. Furthermore, it is possible to design an automatic conversion program for knowledge so that the same judgment and drawings as those skilled in the art can be understood.

Furthermore, using a general-purpose information model that expresses data, diagrams, tables, formulas, rules, and procedures in common, an integrated knowledge base is constructed, and using this, a process plant such as a petrochemical plant, power plant, etc. It can be used for life cycle support such as maintenance, disaster prevention, etc. related to life cycle such as civil engineering, water supply, and building equipment.

According to the present invention, quick analysis of knowledge information can be performed appropriately and in a short time. In other words, it aims to provide answers to both qualitative and quantitative performance, performance, and behavior for inquiries. Therefore, it is a platform that not only axioms and proofs but also procedural processes, calculations and simulations to quickly accumulate and provide knowledge and functions. It also supports functions such as drawing and chart knowledge.

In the knowledge processing layer, a hybrid (hybrid) inference engine that can use a logic language (PROLOG) and a procedural language (JAVA) is implemented. As a result, registered procedures and calculations can also be called from a logic language and executed as part of knowledge processing. Conversely, knowledge processing can be performed by calling a deductive reasoning function as part of a general procedural program. Accordingly, since a language suitable for the contents of processing can be used, it is easy to develop an application system suitable for the purpose, and to gradually improve performance.

Furthermore, according to the present invention, there are provided means for clarifying various forms of knowledge in various areas including tacit knowledge, and further providing means for formalization that can be used in a computer system. In addition, it is an object (knowledge object) in a unified knowledge expression format that is converted into formal knowledge and converted into a knowledge base. By using recent XML, knowledge can be used even in a distributed environment on a network. Since it is possible to provide a standard specification that enhances the knowledge, it is possible to integrate knowledge of various fields and forms and support problem solving necessary for human beings.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
“First Embodiment”

In the first embodiment of the present invention, in a knowledge base system comprising a host computer system (111) and input / output and display unit (101) of the host computer (111), the host computer (111) A class library layer (119), a knowledge base layer (117), a knowledge processing layer (115), and a presentation layer (113) are configured. The class library layer (119) includes behavior and properties (elements / attributes). A class library of knowledge objects to be expressed (objects of knowledge expression format) is provided, and the knowledge base layer (117) is used or defined by each element knowledge as an object of knowledge expression format of the class library layer (119). The knowledge processing layer having (a set of) knowledge objects represented 115) is a protocol (procedure) for processing information based on information input from the input / output unit (101) and knowledge stored in the knowledge base layer (117) or the class library layer (119). The presentation layer (113) obtains data necessary for display from the knowledge processing layer (115), and passes information to the input / output unit / display unit (101). 101) is a knowledge base system that displays (images or sounds) or prints an answer to an inquiry based on information in the presentation layer (113).

The basic system as a base may be a so-called client / server type system, a web server type, or a general-purpose machine, but is preferably a platform (computer base system) that handles an object-oriented language.

The class library layer (119) includes a class library of knowledge objects (knowledge expression format objects (described later)) expressed by behavior and properties (attributes / elements). The class library mentioned here contains various representation forms of knowledge.

Knowledge objects here are classified according to their calling method,
Attribute corresponding to value call Association between objects Rule Calculation module Formula, Axiom, Theorem Procedure An object or a list of its attributes.

The knowledge base layer (117) uses or defines element knowledge of knowledge objects in the knowledge representation format of the class library layer (119), and has a knowledge object or a set thereof. In addition, knowledge objects corresponding to items held in the class library layer (119) may be held.

The knowledge processing layer (115) obtains information based on information obtained from the input / output unit / display unit (101) and knowledge in the knowledge base layer (117) or the class library layer (119). Has a protocol (procedure) to process. An answer that does not give a solution (for example, error handling) is included in the form of providing some kind of answer to an inquiry that goes through the presentation layer (113) or an inquiry that enters directly or indirectly.

The presentation layer (113) obtains data necessary for display from the knowledge processing layer (115) and passes information to the input / output unit / display unit (101). The input / output unit / display unit (101) displays (text, image, or sound) or prints an answer to an inquiry based on information in the presentation layer (113), and a person to be accessed is a person. Or computer system.

A conceptual block diagram of the present invention is shown in FIG. As shown in the figure, the configuration of this system is a knowledge base system comprising a host computer system (111) and an input / output / display unit (101) of the host computer. In other words, it is composed of a presentation layer (113), a knowledge processing layer (115), a knowledge base layer (117), and a class library layer (119).

The basic system serving as a base may be a so-called client / server type system, a web server type, or a general-purpose machine, but is preferably a platform (computer base system) that can handle an object-oriented language.

The class library layer (119) has a class library of knowledge objects (knowledge expression format objects) expressed by behavior (or behavior) and attributes (or properties). The class library mentioned here contains the basic expression form of knowledge.

The knowledge base layer (117) includes (a set of) knowledge objects that are used or newly defined and represented by each element knowledge, which is an object in the knowledge representation format of the class library layer (119). In addition, classes belonging to the class set listed in the class library layer (119) may be held. The knowledge here includes knowledge information included in character data, image data, voice data, tabular data, rules, mathematical formulas, axioms, theorems, procedures, and the like that exist in various forms.

The knowledge processing layer (115) has a protocol (procedure) for processing information based on information input from the input / output unit and knowledge in the knowledge base layer (117) or the class library layer (119). Have. It provides some answers to inquiries that go through the presentation layer (113) or directly or indirectly into each layer (115 to 119) of the host computer (111), and does not give a solution (for example, error handling) Etc.) Answers are also included.

The presentation layer (113) obtains data necessary for display from the knowledge processing layer (115), passes information to the input / output / display unit (101), and the input / output / display unit (101) An answer to the inquiry is displayed (text, image, or sound) or printed based on the information of the presentation layer (113), regardless of whether a person to be accessed is a person or a computer system.

  Hereinafter, the knowledge expression format constituting the present invention refers to description in the expression format described in Table 1 below. “Knowledge” is an expression of the knowledge and also an expression of meaning, which is to express all target objects using “association”. That is, “knowledge” can express a model centered on a verb by predicate logic. This represents all events and objects. Based on this theory, in the present invention, the basic expression forms of knowledge classified into the above seven are defined as shown in the following table.

The knowledge expression (notation) described above, that is, an object in the form of knowledge expression, may be hereinafter referred to as a “knowledge object” for convenience.

The above table lists the main ones, but in reality, the library is registered as a number of object classes, and the present invention also has a new registration function. Using the expression formats listed in the above table as basic rules, various objects are converted into knowledge expression formats (knowledge objects). However, the expression format is not limited to the above-mentioned restrictions, but based on the ontology concept, the target fact or thing is expressed by a combination of “behavior” and “property” (or attribute). The table above provides the main rules (or constraints) for that. Therefore, the system according to the present invention realizes the knowledge notation that unifies the expression format which is also a constituent requirement of the present invention by the rules (or restrictions) in the above table, that is, the “knowledge object” representation. .

According to the present invention, quick analysis of knowledge information can be performed appropriately and in a short time. In other words, we have prepared an “inquiry” method that is suitable for “quick analysis of knowledge information”, and it is easy to answer performance and behavior from both “qualitative” and “quantitative” aspects. It is possible to obtain. Therefore, not only “axioms” and “proofs”, but also “procedural processing”, “calculations and simulations” can be used to quickly utilize “knowledge” and “functions” accumulated.

“Embodiments 2 and 3”
The second and third embodiments of the present invention will be described in detail with reference to the drawings.
FIG. 3 shows a conceptual configuration of the knowledge base system of the second embodiment or the third embodiment of the present invention. As shown in the figure, the configuration of this system is a knowledge-based system comprising a host computer system (211) and an input / output / display unit (201) of the host computer. As a typical configuration, it includes a presentation layer (213), a knowledge processing layer (215), a knowledge base layer (217), and a class library layer (219).

  Further, the host computer system (211) accepts an interface part (223), a drawing, a table, and other data (231) for conversion to a knowledge base for accepting data from the external database (221). Comprehension system (235) composed of digital data comprehension / table comprehension and knowledge representation format conversion unit (233), and external calculation function (241) for providing necessary knowledge information . Connection to an external business system (251) is not necessarily required, but is shown for explanation.

  The external database (221), the understanding of drawings, tables, etc. and the understanding system (235) of the knowledge representation format conversion unit (233), the external calculation function (241) are not all necessary, but one of them, A plurality of combinations are connected to the host computer system (211). Here, the analog data may be a hard copy of a drawing or a table. In this case, the means for converting to digital data is diverted, and a hard copy such as paper is tagged and registered as knowledge information. It may be a means to do. The understanding system unit (233) can be provided with a means for diverting such tagging means and extracting knowledge. These also extend the knowledge base system functions of the knowledge processing layer (115 or 215), knowledge base (117 or 217) and class library layer (119 or 219) of the host computer described in FIG. 1 or FIG. It is something to strengthen.

The external business system (251) is a general-purpose machine (so-called mainframe), a client / server type system, or a WEB system used in a basic business. The same applies to the external database (221).
The conversion from the external database (221) to the class library layer (219) to the knowledge representation object (knowledge object) is performed by the knowledge representation format conversion unit (223). The conversion unit (223)
Attributes corresponding to value (or variable) calls Associations between objects Rules Calculation module Formulas, axioms, theorems Procedural processing, and event processing Depending on the list of objects or their attributes, conversion is performed based on the rules in the preceding table . For details, the following operations are performed.

  The knowledge representation format conversion unit (225) from the external database (221) to the class library layer (219) uses the element knowledge of the knowledge representation format object of the class library layer (219) to newly define the knowledge base layer. (217). The knowledge base layer (217) has knowledge objects (or collections thereof). In addition, classes belonging to classes (or class sets) listed in the class library layer (219) may be held. The knowledge object mentioned here refers to information included in character data, image data, voice data, tabular data, rules, mathematical formulas, axioms, theorems, and procedures existing in various formats. It is defined by an object notation conforming to the knowledge expression format [Table 1] defined by the invention.

  This knowledge (or knowledge object) can be expressed by the object or the relationship between objects based on the object-oriented concept or ontology concept expressed by behavior (behavior) and its attributes (properties). An “association” between objects is also an object.

The knowledge processing layer (215) has a protocol (procedure) for processing information based on information input from the input / output unit and knowledge in the knowledge base layer (217) or the class library layer (219). Have. Some kind of answer is provided to inquiries via the presentation layer (213) and inquiries directly or indirectly entering each layer (215 to 219) of the host computer (211). Answers that do not provide a solution (eg, error handling) are also included.

  The presentation layer (213) obtains data necessary for display from the knowledge processing layer (215), and passes information to the input / output / display unit (201). In other words, the input / output / display unit (201) displays or prints an answer to an inquiry from a user or a system operator as text or an image (or sound) based on information in the presentation layer (213). Regardless of whether the inquiry source is a person or a computer system.

“Fourth, fifth and sixth embodiments”
The fourth, fifth, and sixth embodiments of the present invention will be described in detail with reference to the drawings.
FIG. 4 is a conceptual configuration diagram of the knowledge base system according to the fourth, fifth, or sixth embodiment of the present invention.

As described with reference to FIGS. 1 to 3, information such as diagrams, tables, mathematical formulas, rules, and procedures is performed through the knowledge representation format conversion unit (323) or (333). A file of an object class such as a drawing or a table converted into a knowledge object is stored in the knowledge base layer (317) and is called up as necessary. Similarly, external data is also stored in the knowledge base layer (317) as a knowledge object in advance through the knowledge representation format conversion unit (323) and is called up as necessary. Alternatively, it is registered as a knowledge object in advance from the external calculation function (341), and is called up and used by the knowledge processing layer as necessary. At this time, a remote procedure call in a procedural language is used. Note that the mechanism of conversion into knowledge objects such as diagrams, tables, and calculation modules described above has already been described in detail in the first to third embodiments, and is omitted here.

  The fourth embodiment is characterized in that the knowledge processing layer (315) is configured by using a procedural language as a logical language. As shown in Table 1, the basic expression form of knowledge can be realized by a procedural language. In the present invention, knowledge needs to be expressed as an object, and at the same time, it needs to be expressed as an “association” between objects, which is often expressed predicate logically, and it is better to express knowledge objects in a logical language. Often simplified. Therefore, as a feature of the present invention, in order to take advantage of both advantages, a combined method has been developed.

The parts corresponding to the fifth example and the sixth example are respectively the external calculation function (341) and the logic of the knowledge processing layer (315) of this embodiment for processing information from the external business system. A combination of type functions and procedural calculation functions may be used. As a result, functions or knowledge object formats suitable for the logic language are described in advance based on the logic language and stored in advance in the knowledge base. In addition, if a function that is good at a procedural language is described as a knowledge object, a description based on the language can be made and the knowledge base can be saved. By using these in combination, high-speed inference operations can be performed. This function is mainly used when the accumulated knowledge is used after the knowledge base is constructed. That is, when the user of the knowledge base according to the present invention makes a “knowledge” inquiry to the host computer (311) constituting the knowledge base using the input / output unit / display unit (301), the answer is: This mechanism is used. Inquiries about this knowledge base are usually used with a guide provided on the display screen, such as an eye milk format, so the user is unaware of whether it is a logical language or a procedural language. .
In addition, this is particularly convenient when inquiring about a problem location in an electronic design drawing or using registered design information.

  Further, the host computer system (211) accepts an interface part (223), a drawing, a table, and other data (231) for conversion to a knowledge base for accepting data from the external database (221). Comprehension system (235) composed of digital data comprehension / table comprehension and knowledge representation format conversion unit (233), and external calculation function (241) for providing necessary knowledge information . Connection to an external business system (251) is not necessarily required, but is shown for explanation.

  The external database (321), the understanding of drawings, tables, etc. and the understanding system (335) of the knowledge expression format conversion unit (333) and the external calculation function (341) are not all necessary at the same time, but one of them. A plurality of combinations are connected to the host computer system (311).

  In the knowledge processing layer, an inference engine using both a logic language (eg, PROLOG) and a procedural language (eg, JAVA) was developed and implemented. As a result, it is possible to call a registered procedure or calculation from a logic language and execute it as a part of knowledge processing. Conversely, knowledge processing can be performed by calling a deductive reasoning function as part of a general procedural program. Accordingly, since a language suitable for the contents of processing can be used, it is easy to develop an application system suitable for the purpose, and to gradually improve performance.

“Seventh, eighth, ninth, tenth and eleventh embodiments”
The seventh embodiment, the eighth embodiment, the ninth embodiment, the tenth embodiment, and the eleventh embodiment of the present invention will be described in detail with reference to the drawings.
FIG. 5 is an explanatory diagram of the seventh, eighth, ninth, tenth or eleventh embodiment of the present invention, and shows the conceptual configuration of the knowledge base system.

Similar to the description in FIG. 1 to FIG. 3 or FIG. 4 described above, information (431) such as figures, tables, legends, symbols, design standards, procedures, etc. is performed through the drawing / table etc. understanding system (435). . This understanding system (435) for drawings and tables has the same function as the knowledge expression format conversion unit in FIG. 3 or FIG. 4, but in the seventh, eighth and ninth embodiments, The mechanism of converting data such as drawings and tables into knowledge representation objects (knowledge objects) by the drawing / table understanding system (433) and storing them in the knowledge base layer will be described in detail. . In addition, knowledge objects are reused as needed or called to be used for analysis.
A knowledge object is stored in the knowledge base layer (417) from the external database (421) through the knowledge representation format conversion unit (423), which will be described in detail later. From the external calculation function (441), a remote procedure call in a logical language or a procedural language calls a calculation module or the like prepared as necessary. This is already shown in FIG. 3 or FIG. Similar to what is described. Therefore, detailed description is omitted here.

  FIG. 6 simply shows the knowledge process such as diagrams, tables, mathematical formulas, functions, rules, procedures, and the like. The following explains how information such as diagrams, tables, functions, rules, procedures, etc. is incorporated into the knowledge base. Individual explanations follow.

  FIG. 8 is a flowchart showing an outline of the drawing knowledge process. This is because the computer system recognizes various information contained in the drawing, such as symbols, symbols, legends, connection relationships of symbols and symbols, or the characteristics, components, standards, etc. of these items semi-automatically. , Taking it as knowledge and accumulating it as a knowledge base. The accumulated knowledge base is reused. It also means that the equipment, facilities, joints, and connection relationships entered in the drawings in the target environment being used are also used for equipment state analysis and the like.

Here, a process from recognition of a design drawing to incorporation of the drawing into a knowledge base is shown. This is done with the support of a computer system, and here it will be called "an understanding system for drawings and tables".
The procedure is as follows.
Drawing diagnosis Drawing correction Conversion to description in XML format Tuning Symbol recognition
The process procedure of final drawing comprehension is completed by repeating part of the process beyond topology recognition. In other words, the process of understanding drawings
Clarification of the intention of the drawing creator --- To obtain necessary information and knowledge from the knowledge base of the knowledge drawing, and to obtain an inquiry and answer.
Also used for repeated diagnosis of knowledge-based drawings.
It is.

In addition, the following points are prepared for using the knowledge-based drawing.
Device search function Route search function Range search function Comprehensive search with related data

The flowchart of FIG. 8 will be described below.
(1) Drawing diagnosis (S601)
Since there is a lot of arbitraryness in how to read the implicit intentions of the creator of the drawing, there is a risk of misreading the entire picture, so there is a risk of searching for and correcting logical mistakes in the drawing, and component pixel Perform blocking etc. This is an operation to remove obvious errors and logical contradictions to the extent possible. This is step S601 in FIG.
(2) Primitive recognition In the flowchart of the drawing understanding process in FIG. 6, it is considered that primitive recognition is included in the drawing diagnosis part. This indicates a minute configuration that constitutes a symbol before being recognized as a symbol. It can also be said to be a unit constituting a symbol.
(3) XML conversion (S603)
A computer-aided design (CAD) system captures drawing data output in, for example, DXF format, but intelligent CAD drawings that can be exported in DXF format are automatically converted by an interpreter. Many computer aided design (CAD) systems can output, for example, in DXF format. The conversion from the DXF format to the XML to the system according to the present invention is performed.
(4) Tuning (S604)
Next, parameters are set such as setting an allowable error for primitive elements that are not recognized by symbol recognition and topology recognition described in the steps and subsequent steps, and parts where drawings are not recognized.
(5) Symbol recognition (S605)
Symbol shape recognition, which is defined as a set of primitives connected to a pipe. Symbol recognition extracts a symbol class, a relationship between symbols, a symbol state, an annotation, and the like. For piping parts having simple shapes such as valves and flanges, almost all parts are automatically recognized by extracting feature values such as coordinate points and geometric shapes. Symbol recognition is an important step for giving meaning to a knowledge database generated as a collection of pixel data. Thereby, a specific primitive set is defined as a device or a piping component.
(6) Topology recognition (S606)
The entire topology is constructed with lines connecting the symbols, and the intentions and functions of the designer included in the original design drawing are reproduced.
(7) Recognition and registration of legend (legend) (S606 or S607)
Recognition and registration are performed as necessary in the process of S606 or S607.
When a parts dictionary for each drawing (for each drawing creator) called “legend (legend)” is prepared at the time of parts recognition, collation with the information is performed. For example, in plants such as petrochemicals and electric power, many of the devices such as drums and heat exchangers are blocked for each unit, and the information is used as recognition means. When a legend (legend) does not exist, a function for registering a legend is prepared by designating a primitive (also called a set of elements such as points and lines) from an electronic drawing. As described above, when the drawing understanding process proceeds and knowledge representation is possible, the drawing understanding step is started. (S607)
(8) Drawing understanding (S608)
Drawing comprehension is the clarification of the intention and knowledge of the creator included in the drawing. The relationship and function of another device related to the device here but distant are specified, and various information on the route can be immediately provided.
A description will be given below of how knowledge is expressed as a description including knowledge information, that is, an object in a knowledge representation format.

FIG. 8 is a diagram for explaining the outline of knowledge generation while keeping the knowledge classification process classification of FIG. 7 in mind.
In particular, in the design drawings and the like, the objects to be knowledged are items such as equipment and facilities, such as piping connecting the devices, and these items are recognized as symbols. Further, the devices are connected by piping or the like. Furthermore, there is a functional meaning connection in connection with equipment and facilities.

  Furthermore, there are standards, standards, design standards, etc., for these devices, facilities, and the environment in which they are placed. Furthermore, there are also rules, procedures, and numerical calculations to be used when performing process design and system design using these. Information for verification and evaluation tests is also necessary. These items are roughly classified in FIG. All of these items can be captured as objects, and can be expressed in the form of knowledge expression as described above. That is, it can be described as a knowledge object.

  FIG. 9 shows an outline of a drawing knowledge process using actual computer-aided design (CAD) to explain the drawing knowledge process. In FIG. 9, for example, in the plant design drawing, the state of piping between the devices is expressed in a knowledge expression format. That is, each device is described as an object, and the connection between devices is also expressed as an object. As described above, the object is expressed as first-order predicate logic using a logic language suitable for logic operation.

  FIG. 10 shows that port 1 of the pump 101 and port 4 of the drum 201 are connected as a functional expression. In the lower part of FIG. 10, an example of describing the knowledge expression expressed in the present invention is shown as an expression.

  11 has been developed mainly as a representation of three-dimensional piping that is important in design and construction as a physical representation, but there are various methods such as flange, screw, and welding for connecting the physical pipe and equipment. The connection method is also important. Therefore, an example of expression using the structural description method is described in the lower part of FIG.

  FIG. 12 shows that symbols cannot be recognized during primitive recognition in the drawing knowledge process. Instead, each element can be recognized as a primitive.

  FIG. 13 shows an example of a symbol recognition target portion of a design drawing by a computer-aided design (CAD) system in the outline of the drawing knowledge procedure explained in FIG.

  FIG. 14 shows how the symbol of interest on the drawing of FIG. 13 is described as a knowledge representation format, ie, a knowledge object. Here, it is shown how a symbol (symbol 15) representing a valve (Valve) is represented as an object in a knowledge expression format (ie, a knowledge object. In this example, a symbol class name, a specification name, Various information related to this valve such as number, set, port, property, connection information, measuring instrument number, etc. is held as a knowledge object.These attribute or property information is defined in the field of use etc. as necessary. Specifically, the class name of this symbol is “symbol5”, which has an attribute of “valve.” Furthermore, symbol15 is “GateValve” as a specification, and has “6” as the value of a primitive element. Sri_1 as the classification of the hierarchical class It has set information indicating that it is gri_1 below 19 and is hierarchically below.The value of trimmed_line is 1, followed by port information and property information.symbol 6001 is a symbol 1 It is shown that port number 1 is connected to port number 2 of symbol 4. Lastly, the measuring device number (measuring_instrument_number) is symbol 4, which means that there is reference information corresponding to TI 5559. Thus, the knowledge of the symbol of the specific part as shown in FIG.

FIG. 15 explains how to express a flow sheet in a knowledge expression format in the drawing knowledge procedure. If this is a description of a flow sheet in the form of knowledge representation, it is described as shown in FIG. That is, the relationship between the gate valve 6 (V-1021), the tank (T-101), and the pump (P-202) in the portion surrounded by a red frame in FIG. 15 is shown in this example of the embodiment according to the present invention. Shows the knowledge representation format. Actually, if this flow sheet diagram is written in the form of knowledge representation, that is, it is as shown in FIG. This is also recognized as a class file in the same manner as the representation of the knowledge representation of the symbol representation (the knowledge representation format written out as a result of symbol recognition).

FIG. 17 shows an expression format when symbols, symbol classes, relationships between symbols, symbol states, annotations, and the like are extracted from the outline of the drawing knowledge procedure.
Example) Line connection relationship (A, B, C, D, E)
A: Line ID
B: Symbol ID (connection source)
C: Port ID (connection source)
D: Symbol ID (destination)
E: Port No. (access point)

  FIG. 18 shows an example of the outline (9) -legend (legend) of the drawing knowledge process. Although not described here, the legend (legend) is registered as a prototype of the class. For example, the symbol class in FIG. 13 shows the knowledge object representation of symbol5, but if it is “prototype” instead of “instance”, this means a legend (legend). Thus, the embodiment of the present invention is designed based on the object-oriented concept. In other words, the legend corresponds to a prototype in terms of object orientation.

  FIG. 19 shows that symbols of various valves such as a gate valve, a globe valve, and a check valve are held as knowledge information. In this example, the symbols are recognized, and a list of symbols is displayed on the right side. It is displayed as a tree structure file list.

FIG. 20 shows how the knowledged representation of the material specification is performed. In the outline of the drawing knowledge process, an example is shown which shows how to make a material specification knowledge. Here, the material specifications of “piping class” are shown in a table format. In this way, the information shown in tabular form is
As table data identification information, class name,
attribute,
By giving an attribute value or the like, for example, information corresponding to knowledge included in a table can be obtained.

The meaning of the table column
The object identifier, attribute identifier, and value are expressed in a standard format. Many tables contain knowledge of what is essentially a rule, even if it is not in rule form.

In FIG. 21, in the process of making the equipment specifications into knowledge (knowledge of engineering specifications-equipment specifications, material selection criteria, etc.), knowledge of engineering specifications-equipment specifications, material selection criteria, etc. are prepared. Further details are described in FIG. The notation format is as follows. That is, the piping material selection criteria is
・ Piping parts types: valves, pipes, flanges, etc. ・ Piping size: 2B, 3B, etc. ・ Piping class: Furthermore, use of fluid flowing in piping, operating temperature, operating pressure, etc. It will be decided if conditions are decided.
If P1, S1, etc. are decided,
・ Pipe parts specifications: Model, material, pressure rating, etc. are all determined. In other words, it represents a rule that once the specified type, size, and piping class specify the range of fluid, operating temperature, and pressure, piping parts other than the specified specifications must not be used.
Class name (object identifier, attribute identifier, its value)
It is expressed in a standard format.

22 and 23, in the process of making knowledge of equipment specifications (knowledge of engineering specifications, equipment specifications, material selection criteria, etc.), knowledge of engineering specifications-equipment specifications, material selection criteria, etc. are prepared. Further details are displayed. In these lists, the notation format of the knowledge representation format is as follows.
Knowledge-representation formats such as equipment specifications and material selection criteria can be described as:
stream_property (Stream, Attr, Value).
Stream object identifier
Attr attribute identifier
Value value

  24 and 25 show examples of device specification knowledge. This means that the information in the table is recognized as knowledge information and written in a knowledge representation format to obtain knowledge information. That is, in the example of FIG. 24, it is shown that the item_no of “equipment” is specified by E-523 and described by the notation method shown in FIG.

FIG. 25 illustrates a method of making a procedure knowledge.
The knowledge representation of the procedure may be considered as the relationship between the initial state object and the completed state object. There are many cases in which procedural processing such as display on a graphic screen is used as part of knowledge processing. In particular, in a complex system such as a plant, it is very effective to color-display and display objects logically searched on a flow sheet.
In other words, the procedure can be said to arrange predicates. However, it is more efficient to use a dedicated language and program suitable for procedural processing such as display and user interface, including numerical calculation, rather than directly realizing the procedure with an axiom and proof system.
display (X, init, end).
This means that the procedure is called by the procedure name display, the target object X, and the initial state object init to call the procedure, and when the procedure is completed, the completion state end is set.

FIG. 26 illustrates a point handled by using numerical calculation as a knowledge base.
At D-201, feed (FEED) is calculated by vapor-liquid equilibrium separation calculation, and the separated overhead gas fraction (OVD) and bottom liquid fraction (BTM) are calculated.
In this example, the above calculation can be calculated using the following equation.
equilibrium_separation ('D-201, FEED, OVD, BTM).

It is the same as the calling format of a general module. When called, the calculation is performed and the result is returned. Similar to procedures, numerical computations can be treated as relations by considering them as object relationships between input variable objects and output variable objects. That is, an API that registers as a calculation module having an argument and calls the calculation module is handled as a formal knowledge expression.

  FIG. 27 illustrates a mechanism for utilizing the knowledge base knowledge (knowledge object) constructed in addition to the process of knowledgeization illustrated in FIG. 6 and described in FIGS. 7 to 26 thereafter. Knowledge is used by combining questions and answers.

  According to the present invention, a knowledge base described by knowledge notation such as diagrams, tables, mathematical formulas, functions, rules, procedures, standards, standards, etc. A process for retrieving and reusing the knowledge object will be described below with reference to an embodiment. Knowledge is defined as an object, and the object enables a composite method that can be called and used by a procedural language or a logical language.

Here, the question is given to the knowledge base system according to the present invention by the description of the predicate logic type, and the knowledge base system searches the object existing in the knowledge base for the given question, and converts it into necessary attributes and attributes. The relevant knowledge object class is searched with the relevant value, and the class is instantiated by passing the value, and returned as an instance object of the answer to the question. This may be accompanied by the behavior that the object invokes, including invoking the required action by invoking the associated object.

Hereinafter, an actual example of knowledge call will be described. In the following example, in the first to ninth embodiments in the knowledge base system according to the present invention, questions are registered in a predicate format. In the example shown in FIG. 28, the description of the question for obtaining the flow velocity of the pipe is shown. According to the formula of this question, “pipe size from pipe number as pipe identifier, stream number corresponding to pipe number is given, as a result, flow rate is obtained, pipe inner diameter is searched from JIS industrial standard, and the accompanying“ flow velocity It answers the flow velocity from The search formula in this example is expressed by the following formula.

line_velocity (A, D, F, V):-
line_no (A, line_id, B),
line_no (A, line_size, E),
stream_property (C, line_id, B),
stream_property (C, flow_rate_v, D),
jis_pipe_size (C, 'sch_5c', pipe_id, F),
V is (D / 3600) / (F / 1000/2) / (F / 1000/2) /3.14.

The calculation module is created in a procedural language when speeding up the calculation, and the call format is registered as knowledge / present and is used for calculation. As described above, the seventh embodiment, the eighth embodiment, the ninth embodiment, the tenth embodiment, and the eleventh embodiment of the present invention are described above with reference to the external database system (425). The embodiment is realized by combining the figure / table understanding system (435), the external calculation function (441), or the external business system (451). This will be described in detail.

“Embodiment 12”
A twelfth embodiment of the present invention will be described in detail with reference to the drawings.
FIG. 29 shows a twelfth embodiment of the knowledge base system according to the present invention, and includes the first to eleventh embodiments of the present invention therein. Details of FIG. 29 will be described below. The host computer system (711) and the host computer (711) are a knowledge base system in which an input / output and display unit (701) is connected via a network. The host computer (711) It is composed of a library layer (719), a knowledge base layer (717), a knowledge processing layer (715), and a presentation layer (713). The class library layer (719) is expressed by behavior and properties (elements / attributes). The knowledge base layer (717) uses or defines each element knowledge that is a knowledge representation object of the class library layer (719). A knowledge object to be represented The layer (715) is a protocol (procedure) for processing information based on information input from the input / output unit (701) and knowledge stored in the knowledge base layer (717) or the class library layer (719). The presentation layer (713) obtains data necessary for display from the knowledge processing layer (715), passes information to the input / output unit / display unit (701), and inputs / outputs the display / input unit / display. The unit (701) is a knowledge base system that displays or prints the answer to the inquiry based on the information in the presentation layer (713) as characters, images, or sounds.

The knowledge base composed of the host computer (711), the input / output and display unit (701), and the network unit (723 composed of 725, 727, and 729) connecting them further includes an external application. One or a plurality of units (705), computer-aided design system (CAD) (707), drawing understanding system (709), and analysis tool (711) may be connected. Here, the external application unit (705), computer aided design (CAD) system (707), drawing understanding system (709), analysis tool (711), application interface (721), and network unit, (723) )so. Connected to the knowledge base.

The present embodiment enables a configuration in a so-called distributed environment via a network, and realizes an environment in which a remote user uses the knowledge base system according to the present invention. A knowledge object is stored in the knowledge base layer (417) from the external database (421) through the knowledge representation format conversion unit (423), which will be described in detail later.
From the external calculation function (441), a remote procedure call in a logical language or a procedural language calls a calculation module or the like prepared as necessary. This is already shown in FIG. 3 or FIG. Similar to what is described. Therefore, detailed description is omitted here.

FIG. 30 illustrates an embodiment according to the present invention.
A user inputs information necessary for making a knowledge base into a flow sheet (811), a specification sheet (813), a calculation formula (815), and a rule (817) in advance, and makes it knowledgeable. Here, assuming that a knowledge base system according to the present invention is used on the premise of a petrochemical plant or a power plant, design (821), safety management (823), operation management (825), facility management ( 827) and an example of use in environment management (829).

FIG. 31 shows an embodiment in which the above-described plant data, that is, the plan and the overall design information, equipment list, material information, industry standard, standard, and design standard are converted into a knowledge base to constitute a knowledge base according to the present invention. The main functions will be described below.
(1) Device search function When a device (valve, reducer, etc.) name included in the drawing is selected, the corresponding device in the drawing is searched and colored.
(2) Route search function A route from an arbitrary device in the drawing to another arbitrary device is searched and colored.
(3) Range search function The entire range from the arbitrary device in the drawing to the termination condition is searched and displayed in color.
(4) Comprehensive search with related data Using knowledge such as equipment number and piping number recognized from drawings, search and simulation are performed in cooperation with related data other than drawings (operating procedure, risk location prediction, etc.).

  FIGS. 32 and 33 show the results of the above-described range search performed on the above-described knowledge base system in a process plant such as a petrochemical and power plant. FIG. An example of the search result is shown, and FIG. 33 is an enlarged view thereof. In this way, the user can ask related questions to the knowledge base and obtain necessary design information, operation management information, safety information, facility management information, and environmental management information.

Conceptual diagram of knowledge base system according to the present invention Knowledge representation rules for elements Conceptual diagram of the first, second or third embodiment of the knowledge base system according to the present invention Conceptual diagram of a fourth, fifth or sixth embodiment of the knowledge base system according to the present invention. Conceptual diagram of a seventh, eighth, ninth, tenth or eleventh embodiment of a knowledge base system according to the invention Knowledge process of diagrams, tables, formulas, functions, rules, procedures, etc. Knowledge process of diagrams, tables, formulas, functions, rules, procedures, etc. Drawing Knowledge Making Process-Outline of Drawing Knowledge Making Procedure (1) Drawing knowledge process --- Outline of drawing knowledge process (2) Drawing knowledge process-functional expression Drawing knowledge process-physical representation Drawing knowledge process-Recognition of CAD drawing symbols Drawing knowledge process-Recognition of CAD drawing symbols Drawing knowledge process-flow sheet diagram Process of drawing knowledge-Knowledge representation of flow sheet diagram Drawing knowledge process-list of recognized symbols Knowledge of material specifications Knowledge of equipment specifications Engineering specifications, equipment specifications, material selection criteria, etc. Knowledge of equipment specifications Engineering specifications, equipment specifications, material selection criteria, etc. Knowledge of equipment specifications (1) Knowledge of equipment specifications (2) Explanatory drawing of knowledge calculation of numerical calculation Process of making knowledge of diagrams, tables, formulas, functions, rules, procedures, etc., and search call (use) process Conceptual diagram of a twelfth embodiment of the knowledge base system according to the present invention. Conceptual diagram of a twelfth embodiment of the knowledge base system according to the present invention. Knowledge of plant data Range search results Enlarged view of range search results

Explanation of symbols

101 I / O and Display Unit 111 Host Computer 113 Presentation Layer 115 Knowledge Processing Layer 117 Knowledge Base Layer 119 Class Library Layer 201 I / O and Display Unit 211 Host Computer 213 Presentation Layer 215 Knowledge Processing Layer 217 Knowledge Base Layer 219 Class Library Layer 221 External Database 221 Knowledge representation format conversion unit 231 Drawing / table etc. 233 Knowledge representation format conversion unit 241 for drawing / table etc. External calculation function 251 Business system 301 Input / output and display unit 311 Host computer 313 Presentation layer 315 Knowledge processing layer 317 Knowledge base layer 319 Class library layer 321 External database 321 Knowledge representation format conversion unit 331 Drawing / table etc. 333 Drawing / table etc. understanding system and knowledge representation format conversion unit 34 1 External calculation function 351 Business system 401 Input / output and display unit 411 Host computer 413 Presentation layer 415 Knowledge processing layer 417 Knowledge base layer 419 Class library layer 421 External database 421 Knowledge representation format conversion unit XML format 431 Drawing / table etc. 433 Table understanding system and knowledge representation format conversion unit 441 External calculation function 451 Business system 511 Diagram format data 513 Table format data 515 Function format data 517 Rule 518 Procedure 503 Automatic knowledge conversion 505 Integrated knowledge base system 507 Question / answer 509 Application / Question / Answer S601 Drawing diagnosis diagram format data S602 Drawing correction S603 XML conversion S604 Tuning S605 Symbol recognition S606 Topology recognition S607 Conditional branching S608 Drawing understanding 01 input and output, and a display unit 703 external application 705 computer-aided design system (CAD)
707 Drawing understanding system 709 Analysis tool 711 Host computer 713 Presentation layer 715 Knowledge processing layer 717 Knowledge base layer 719 Class library layer 721 Application interface part (adapter)
723 Network unit 725 Network unit 727 Network unit 729 Network unit

Claims (12)

In a knowledge base system comprising a host computer and an input / output unit and / or a display unit connected to the host computer,
The host computer has a class library layer, a knowledge base layer, a knowledge processing layer, and a presentation layer,
The class library layer includes a class library of objects (knowledge objects) in a knowledge representation format expressed by behavior and properties (elements / attributes).
The knowledge base layer has a set of knowledge objects that are used or newly defined and represented by each element knowledge of knowledge expression format objects (knowledge objects) of the class library layer,
The knowledge processing layer has a protocol (procedure) for processing information based on information input from the input / output unit and knowledge in the knowledge base layer or the class library layer,
The presentation layer obtains data necessary for display on the display unit from the knowledge processing layer, and passes information to the input / output unit.
The input / output unit displays an answer to an inquiry based on information in the presentation layer (outputs as text, image, sound, or print information), and is an integrated knowledge base system. The class library layer or the knowledge base layer is connected to an interface unit with an external database,
The interface unit includes conversion means for converting information from the external database into an object (knowledge object) in a knowledge representation format,
2. The integrated knowledge base system according to claim 1, further comprising conversion information storage means for storing the conversion result of the information (object conversion information in a knowledge expression format) in the class library layer or the knowledge base layer. The knowledge base layer converts analog data information / digital data information such as drawings, diagrams, tables, text data, audio data, or other analog data information / digital data information into an object (knowledge object) in a knowledge expression format. Conversion means for
3. The conversion information storage means for storing the conversion result of each piece of information (that is, the conversion result to an object (knowledge object) in a knowledge expression format) in the knowledge base layer. Integrated knowledge base system. In a knowledge base system comprising a host computer and an input / output unit and / or a display unit connected to the host computer,
The host computer has a class library layer, a knowledge base layer, a knowledge processing layer, and a presentation layer,
The class library layer has a class library of objects (knowledge objects) expressed in behavior and properties (elements / attributes).
The knowledge base layer has a set of knowledge objects that use or newly define and represent each element knowledge that is an object (knowledge object) in the knowledge representation format of the class library layer,
The knowledge processing layer has a protocol (procedure) for processing information based on information input from the input / output unit and knowledge in the knowledge base layer or the class library layer,
The presentation layer obtains data necessary for display on the display unit from the knowledge processing layer, and passes information to the input / output unit.
The input / output unit displays an answer to the inquiry based on the information of the presentation layer (output as text, image, audio or print information),
The integrated knowledge base system, wherein the knowledge processing layer includes inference means configured by a procedural language and a logical language.   The integrated knowledge processing layer according to claim 1, 2, 3, or 4, wherein the knowledge processing layer includes an inference means having a function of calling an external calculation function and using it for a logical operation or a procedural operation. Knowledge base system.   6. The integrated knowledge base system according to claim 1, 2, 3, 4, or 5, wherein the knowledge processing layer and / or the presentation layer includes an interface unit with an external business system. In a knowledge base system comprising a host computer and an input / output unit and / or a display unit connected to the host computer,
The host computer has a class library layer, a knowledge base layer, a knowledge processing layer, and a presentation layer,
The class library layer includes a class library of objects (knowledge objects) in a knowledge representation format expressed by behavior and properties (elements / attributes).
The knowledge base layer has a set of knowledge objects that use or newly define and represent each element knowledge that is an object (knowledge object) in the knowledge representation format of the class library layer,
The knowledge processing layer has a protocol (procedure) for processing information based on information input from the input / output unit and knowledge in the knowledge base layer or the class library layer,
The presentation layer obtains data necessary for display on the display unit from the knowledge processing layer, and passes information to the input / output unit.
The input / output unit displays an answer to the inquiry based on the information of the presentation layer (output as text, image, audio, or print information), and the interface unit connected to the class library layer or the knowledge base layer includes The second knowledge object conversion means for converting the information from the external database into the XML format by the first knowledge object conversion means for converting the information into the knowledge expression format object and converting it into the knowledge expression object (knowledge object). An integrated knowledge base system characterized by comprising: It has a conversion means to convert information into knowledge representation object (knowledge object),
Electronic drawings, diagrams, tables, symbols, legends, formulas, functions, procedures, rules, design standards, standards, or industry standards diagrams, tables, symbols, legends, formulas, functions, procedures, rules, standards, etc. Data input / output means;
A user interface that interactively performs a function of converting the data into an object (knowledge object) in a knowledge representation format;
8. The integrated knowledge base system according to claim 1, 2, 3, 4, 5, 6 or 7. Electronic drawings, diagrams, tables, symbols, legends, formulas, functions, procedures, rules, design standards, standards, or industry standards drawings, diagrams, tables, symbols, legends, formulas, functions, procedures, rules, Input / output means for electronic data such as standards;
Symbol recognition means for recognizing the digitized data as a symbol;
Display means for displaying the digitized data;
A user interface that interactively performs a function of converting the digitized data into a knowledge representation object (knowledge object);
The integrated knowledge base system according to claim 7 or 8, characterized by comprising: Electronic drawings, diagrams, tables, symbols, legends, formulas, functions, procedures, rules, design standards, standards, or industry standards drawings, diagrams, tables, symbols, legends, formulas, functions, procedures, rules, Input / output means for electronic data such as standards;
Symbol recognition means for recognizing the digitized data as a symbol;
Topology relation recognition means for recognizing a relation between a plurality of recognized symbols (joint relation, connection relation, connection relation, connection position relation, etc. as a topology relation;
Means for modifying said symbol recognition means and topology recognition means;
Means for converting recognized symbols and topology elements / attributes into knowledge representation objects (knowledge objects);
The integrated knowledge base system according to claim 7, 8, or 9. Electronic drawings, diagrams, tables, symbols, legends, formulas, functions, procedures, rules, design standards, standards, or industry standards drawings, diagrams, tables, symbols, legends, formulas, functions, procedures, rules, Input / output means for electronic data such as standards;
Symbol recognition means for recognizing the digitized data as a symbol;
Topology relation recognition means for recognizing a relation between a plurality of recognized symbols (joint relation, connection relation, connection relation, connection position relation, etc. as a topology relation;
Means for modifying said symbol recognition means and topology recognition means;
Means for converting recognized symbols and topology elements / attributes into knowledge representation objects (knowledge objects);
11. The integrated knowledge base system according to claim 7, 8, 9, or 10, wherein general drawing information is converted into a knowledge expression object (knowledge object). In a knowledge base system comprising a host computer and an input / output unit and / or a display unit connected to the host computer,
The host computer has a class library layer, a knowledge base layer, a knowledge processing layer, and a presentation layer,
The class library layer includes a class library of objects (knowledge objects) in a knowledge representation format expressed by behavior and properties (elements / attributes).
The knowledge base layer has a set of knowledge objects that are used or defined and expressed by each element knowledge of knowledge expression format objects (knowledge objects) of the class library layer,
The knowledge processing layer has a protocol (procedure) for processing information based on information input from the input / output unit and knowledge in the knowledge base layer or the class library layer,
The presentation layer obtains data necessary for display on the display unit from the knowledge processing layer, and passes information to the input / output unit.
The input / output unit displays an answer to an inquiry based on information in the presentation layer (outputs as an image, audio, or print information), and is an interface unit connected to the class library layer or the knowledge base layer ? Has a second knowledge objectification conversion means for converting the information from the external database into the XML format by the first knowledge objectification conversion means for converting the information from the external database into an object (knowledge object) in the knowledge expression format. An integrated knowledge base system
In a network environment, the component information in which the host computer, the computer design (CAD) system, the drawing understanding system, the external computing system, and the external database are distributed and transferred is transferred in a unified knowledge expression format. An integrated knowledge base system that can be used by remote users by accumulating various knowledge in the knowledge base.

Images