WO2017222410A1

WO2017222410A1 - Method for semantic searching in an object-process data model

Info

Publication number: WO2017222410A1
Application number: PCT/RU2016/000455
Authority: WO
Inventors: Андрей Николаевич АНДРИЧЕНКО
Original assignee: Общество с ограниченной ответственностью "ЭсДиАй Рисёчь"
Priority date: 2016-06-20
Filing date: 2016-07-19
Publication date: 2017-12-28
Also published as: RU2016124318A

Abstract

The present invention relates to computer-based methods of machine production, and more particularly to selecting and determining the compatibility of technical material objects which interact with one another at different stages in their life cycles: design, operation, repair, transportation, recycling,etc. A method for semantically searching in an object-process data model includes the following stages: classifying objects and processes by creating an ontology, wherein, in each category of objects, at least one component of an object is defined such as to include at least an identifier of the component and one or more attributes which characterize the component in question; a link is formed between each component of the object and at least one process in which the component takes part; for each process, all pairs of objects having a link to the process in question are found; a link is formed which determines the rules of interaction for the objects found in the preceding step for each process; the category of a searched object and at least one process in which the searched object takes part are received from an input device; all objects having a link to the aforementioned process are found. The technical result is an increase in the accuracy and speed of searching for a technical object.

Description

METHOD OF SEMANTIC SEARCH IN OBJECT-PROCESS

DATA MODELS

FIELD OF THE INVENTION [0001] The present invention relates to computational methods of machine-building production, namely to methods of semantic search, as well as selection and determination of compatibility of material and technical objects interacting with each other at various stages of their life cycle: design, operation, repair, transportation disposal, etc. BACKGROUND

[0002] Currently, the manufacturing enterprise uses many software and hardware-software complexes that solve various information, organizational and technological problems, namely: CAD, ERP, PLM, etc.

[0003] Currently, the most promising direction of development of MDM systems is the use of semantic technologies, which makes them more effective when working with engineering data.

[0004] The basis of semantic technology is the concept of a semantic network. The use of semantic technologies as part of hardware and software systems allows for a “meaningful” search for objects in which both the parameters of the desired object and the rules for its interaction with other objects take part. The integration of semantic MDM and CAD allows you to increase the level of decision-making automation in the design process of new products.

[0005] The closest analogue of the proposed technical solution is patent application D1: US 20110131247 A1 "Semantic Management Of Enterprise Resources", patent holder: International Business Machines Corporation, published: 02/02/2011 1. This document discloses information about the semantic management of enterprise resources . However, in this patent information source there are no signs of a semantic search on the semantic data structure and the use of ontologies for storing data. SUMMARY OF THE INVENTION

[0006] The technical problem in this technical solution is the management of reference data in industrial enterprises with the introduction of semantic relationships between information objects.

[0007] The technical result achieved by using the present invention is to increase the accuracy and speed of the search for technological objects.

[0008] The indicated technical result is achieved due to the method of semantic search in the object-process data model, it includes the following steps: formulate the classification of objects and processes by creating ontologies, while at least one component is selected in each category of objects from which it consists an object including at least a component identifier and one or more attributes characterizing a given component; form a connection between each component of the object with at least one process in which this component is involved; for each process, all pairs of objects are found that are associated with this process; form a relationship that defines the rules of interaction for the objects found in the previous step for each process; receive from the input device the category of the desired object and at least one process in which the desired object is involved; receive all objects that are associated with the above process.

[0009] In some embodiments, when forming a classification of objects and processes by creating ontologies, where in each category of objects at least one component is distinguished from which the object consists, including at least a component identifier and one or more attributes, characterizing this component, the identifier is an integer value.

[00010] In some embodiments, when forming a connection between each component of an object with at least one process in which this component is involved, the connection is specified in the form of mathematical and logical operations.

[00011] In some embodiments, when forming a relationship between each component of an object with at least one process in which the component is involved, a component of the object is associated with one process.

[00012] In some embodiments, when forming a relationship between each component of an object with at least one process in which the component is involved, the component of the object is associated with several processes. [00013] In some embodiments, when forming a relationship that defines the interaction rules for the found objects for each process, the object may have several internal states — structures, each of which may be associated with one or more processes.

BRIEF DESCRIPTION OF THE DRAWINGS

[00014] The features and advantages of the present invention will become apparent from the following detailed description of the invention and the accompanying drawings, in which:

[00015] Figure 1 shows an example implementation of a technical solution according to a method for semantic search in an object-process data model;

[00016] Figure 2 presents a diagram of the hierarchy of classes included in the classifiers of tools and metal cutting machines. Objects are associated with associative classes that define compatibility rules. Connections between objects are in turn inherited from processes. To determine the compatibility of a pair of objects, it is necessary to verify that the conditions of the set of rules located in the class hierarchy are met.

[00017] In FIG. 3 shows the selection of components of an axial cutting tool;

[00018] In FIG. 4 shows the compatibility of objects, which is determined by the combined compatibility of their components;

DETAILED DESCRIPTION

[00019] This technical solution can be implemented on a computer, in the form of a system or computer-readable medium containing instructions for performing the above method.

[00020] The technical solution may be implemented as a distributed computer system.

[00021] In this solution, a system means a computer system, a computer (electronic computer), CNC (numerical control), PLC (programmable logic controller), computerized control systems and any other devices that can perform a given, well-defined sequence of operations (actions, instructions). [00023] An instruction processing device is understood to mean an electronic unit or an integrated circuit (microprocessor) executing machine instructions (programs).

[00024] An instruction processing device reads and executes machine instructions (programs) from one or more data storage devices. Data storage devices may include, but are not limited to, hard disks (HDDs), flash memory, ROM (read only memory), solid state drives (SSDs), and optical drives.

[00025] A program is a sequence of instructions for execution by a computer control device or an instruction processing device.

[00026] Below will be described the terms and concepts necessary for the implementation of this technical solution.

[00027] CAD or Computer-Aided Design System - an automated system that implements information technology for performing design functions, is an organizational and technical system designed to automate the design process, consisting of personnel and a set of technical, software and other means of automating its activities. Also, the abbreviation CAD is widely used to designate such systems.

[00028] ERP (Enterprise Resource Planning) is an organizational strategy for the integration of production and operations, human resources management, financial management and asset management, focused on the continuous balancing and optimization of enterprise resources through a specialized integrated application software package that provides a common data and process model for all areas of activity. ERP-system - a specific software package that implements the ERP strategy.

[00029] Product Lifecycle Management (PLM) - product lifecycle management technology. Organizational and technical system that provides the management of all information about the product and related processes throughout its entire life cycle, from design and production to decommissioning. Moreover, various complex technical objects (ships and automobiles, airplanes and rockets, computer networks, etc.). Information about the object contained in the PLM system is a digital layout of this object.

[00030] Information technology manufacturing enterprises face the problem of duplication of reference databases from various suppliers: ERP, PLM, CAD, etc. The solution is to separate reference data from application applications, centralize this data and unify terms (thesaurus) based on ontological data models. This is done through the use of a systematic approach to building a single enterprise information space at the reference data level, which allows for the implementation of a single corporate communication language for automated systems or Master Data Management (master data or master data management).

[00031] Master data management, master data management - a set of processes and tools for the continuous determination and management of company master data (including reference).

[00032] Master data is data with the most important information for an enterprise or business information: about customers, products, services, personnel, technologies, materials, and so on. They change relatively rarely and are not transactional.

[00033] There is also a variation of MDM — PIM class systems (MDM for PIM / Product Information Management): Oracle Product Hub, IBM Infosphere MDM Product Hub, SAP MDM. These systems are designed for centralized management of product reference data (Product Information Management). They allow you to provide a uniform, reference presentation of information about the company's products, so that the various information subsystems of the enterprise can exchange information presented in a single format.

[00034] A semantic network is an information model of a subject domain, having the form of a directed graph, whose vertices correspond to objects of a subject domain, and arcs (edges) define the relationship between them. Objects can be concepts, events, properties, processes. Thus, the semantic network is one way of representing knowledge. In a semantic network, the role of vertices is played by the concepts of a knowledge base, and arcs (moreover, directed ones) define relations between them. Thus, the semantic network reflects the semantics of the subject area in the form of concepts and relations. [00035] Ontology is a specification of a certain subject area, which describes many terms, concepts and classes of objects, as well as the relationships between them. The ontology is designed to provide a consistent, unified vocabulary of terms for the interaction of various corporate information systems.

[00036] Previously, before the semantic search, the semantic data structure is formed by forming a classification of objects and processes.

[00037] FIG. 1 is an example embodiment of a technical solution according to a semantic search method in an object-process data model, in which:

[00038] Step 101: forming a classification of objects and processes by creating ontologies, wherein, in each category of objects, at least one component is distinguished from which the object consists, including at least a component identifier and one or more attributes, characterizing this component.

[00039] The simplest example of ontology construction is the allocation in the structure of an axial cutting tool (RI) of the connecting and cutting parts as independent classified objects (components), which allows them to be used to construct descriptions of similar tools, such as: drill, countersink, reamer, end mill and etc. (Fig. 3). Each of these parts has its own unique composition of attributes.

[00040] The identifier of the component may be an integer value.

[00041] The rules for the interaction of these objects in the specified process are set in the form of mathematical and logical operations.

[00042] The rules for the interaction of these objects in the specified process are set in the form of mathematical and logical operations in scripting programming languages.

[00043] This method can be implemented in a hardware-software complex.

[00044] Step 102: form a relationship between each component of the object with at least one process in which the component is involved.

[00045] On the set of constituent parts (components) of an object, stable groups are formed that characterize the object in the context of certain processes. Each group includes the components of the object associated with a particular process. The same component of an object may be associated with one or more processes.

[00046] Thus, semantic structures are created in the semantic network for material objects that are needed in a specialized context (process): repair, operation, relocation, disposal, etc. The context of an object controls the structure of its constituent parts (i.e., ultimately the composition of attributes) and the relationships of this object with other objects during a certain period of its life cycle. In a broad sense, the structure of each object is considered exclusively from the point of view of the activity of a given material object, i.e. those actions - processes in which this object takes part. We can say that the structure of the object is determined by the possible actions with it.

[00047] Step 103: for each process, all pairs of objects are found that are associated with this process.

[00048] Information about the compatibility of objects is entered into the database. Compatibility rules are distributed according to the classification hierarchy up to instances of objects and their components. Rules belong to two types of relationships: object - object, component - component. Each rule must be associated with a specific context (process), i.e. be context-sensitive. In the end, a semantic network is formed that defines the interconnections and compatibility of objects belonging to different categories. The compatibility of two objects in such a network will be determined solely in the context of a particular process by the combined compatibility of their components belonging to this process.

[00049] Step 104: form a relationship that defines the interaction rules for the objects found in the previous step for each process.

[00050] Thus, an object can have several internal states - structures, each of which can be associated with one or more processes. To determine the compatibility of two objects, it is first necessary to determine the process within which they interact, select the appropriate structure of the objects, and only then check the compatibility of a pair of objects based on the compatibility of their components.

[00051] Step 105: receiving from the input device the category of the desired object and at least one process in which the desired object is involved.

[00052] Get the category of the desired object, the process in which the desired object and related objects involved in this process;

[00053] Select all objects in the category of the desired object, at least one component, which is associated with the above process; [00054] Among the objects selected in the previous step, those objects whose components are compatible with the components of adjacent objects based on the compatibility rules in the context of the above process are selected.

[00055] Also, a method for searching for objects in a semantic data structure can be implemented with a modification:

[00056] in a first step, the parameters of the desired object are further obtained, including at least one attribute;

[00057] after performing the basic steps, additionally select those objects that satisfy the parameters of the desired object.

[00058] This method of semantic search can be performed on a semantic data structure generated by the method described above.

[00059] This method can be implemented and be implemented on a hardware-software complex (platform).

[00060] The input from the input device accepts as search criteria: the type of the desired object and its individual parameters (if any), the name of the process, and available adjacent objects that participate with the sought in this process. The process and related objects in this case act as filters - limiters of the search area of the desired object.

[00061] In the classifier to which the desired object belongs, a search is made for those objects that have a group of components associated with the specified process. Among the selected objects, a search is made for those objects that are compatible with adjacent objects submitted as input within the specified process.

The compatibility of objects is determined through the compatibility of their components included in the group of the same name in the context of a given process. Among the remaining objects, a global search is performed according to the specified parameters of the desired object.

[00062] Step 105: receive all objects that are associated with the above process.

[00063] As a result of semantic search, a limited group of objects that have passed two stages of filtering will be selected: a search by compatibility with adjacent objects in the context of a given process and a search by the specified parameters of the desired object. This search mechanism can significantly reduce the scope of the desired solution and increase the accuracy of the search, and also increases the speed of the search.

EXAMPLES OF IMPLEMENTATION [00064] Assume that the following task has been set in engineering: to develop a plan for processing an axial hole in a solid alloy body part. The source data are:

[00065] type, material and overall dimensions of the part;

[00066] type, accuracy, cleanliness and overall dimensions of the processed surface (structural element - FE);

[00067] Moreover, the equipment was initially determined - a vertical-boring machine 2N135. Now you need to determine which cutting tool (RI) the hole will be machined.

[00068] In the RI classifier there are several types of tools that can handle holes: drills, countersinks, reamers, broaches, cutters, end mills. First you need to select the required type of RI. Within this type of drill, for example, there are several hundred types of drills for different materials: for metal, for wood, for concrete, etc.

[00069] Drills are also divided into types: spiral, feather, cannon, etc. Each drill has a different connecting part (shank: conical, cylindrical, hexagonal, etc.), which determines the possibility of installing the drill on a particular type of equipment.

[00070] In this case, only drills for machining hard metal are suitable with the type of connecting part that is compatible with the RI seat on the 2H135 machine. The dimensions of this drill should correspond to the dimensions of the FE, and the overall dimensions of the RI should not exceed the overall dimensions of the working area of the machine, taking into account the case part installed on it. There will be several units of these types of drills in the entire RI catalog. It is necessary that by the proposed method select only these types of drills from the catalog of RI.

[00071] The task is formulated as follows - under given initial conditions (the type of part is a body with known dimensions, the material of the part is hard alloy, the model of the machine is 2H135, the hole with a given diameter D, L and surface quality) select only those types of cutting tools, which are compatible with the listed objects, i.e. to carry out a semantic, “meaningful” search for objects in which both the parameters of the sought-after RI object (the diameter of the RI must correspond to the diameter of the FE) and the rules for its compatibility with other objects: material and machine, will take part.

[00072] One of the steps in solving this problem is the formation of a semantic data structure in the described manner. [00073] To solve this problem it is necessary:

[00074] Build a classification of technological objects: machines, cutting tools, materials, etc., through the creation of ontologies. Provide each object with attributes and attribute them to a specific component of the object (component).

[00075] When describing each object, unified attributes or attribute groups from this list are used. This solves the problem of discrepancies - when the same characteristic in different classification categories of objects can be called differently.

[00076] The construction of a complete list of unified characteristics (machine, tool, etc.), as well as the formation of the structure of the object that determines the occurrence of components (the tool consists of dir. +. Connecting parts + material + ...) and there is a description of the subject areas (ontology) - see Thomas Grubber's definition given above.

[00077] Without building an ontological model of an object, it is impossible to formalize its relationships with other entities, because compatibility rules for two objects are determined by the combined compatibility of their components (Fig. 4).

[00078] Distribute attributes and components by context. The context of an object controls the composition of attributes, the structure of its constituent parts, and the relationships of this object during a certain period of its life cycle. So a technologist, designer, accountant, and supplier should see only those attributes that he needs, i.e. in its context. Accountants, for example, are not interested in the characteristics of the blade and the connecting part of the RI.

[00079] Enter information about the compatibility of various objects in the form of rules, which are a combination of arbitrarily complex mathematical operations and logical conditions. For example:

[00080] Rule of compatibility between an axial cutting tool and a metal cutting machine: The morse cone of the machine seat must be equal to the morse cone of the connecting part of the cutting tool:

[00081] Machine. Seat. Morse cone = Tool. Connecting part. Morse cone, where “Machine” is the object, “Seat” is the component, and “Morse cone” is the attribute.

[00082] b) Rule of compatibility of the workpiece and the lathe: The workpiece can be installed in the working space of the machine if its length is less than the distance between the centers of the machine and the radius of the workpiece is less than the height of the centers of the machine: [00083] (Machine. Workspace. The height of the centers above the bed> 0.5 * Workpiece. Diameter) And (Machine. Workspace. The distance between the centers> Workpiece. Length)

[00084] The rule can be developed in a scripting programming language: VBScript, JavaScpipt using any logical and mathematical operations.

[00085] Only components and attributes of objects "visible" in the context of a given process participate in the rules. Compatibility rules are distributed according to the classification hierarchy up to the constituent parts of objects (Fig. 2, Fig. 5).

[00086] The compatibility of two objects is defined as the compatibility of their components. So, the compatibility of the RI and the metalworking machine is determined by a set of rules at various levels: whether this type of RI is compatible with this type of machine, is the type of machine seat compatible with the connecting part of the RS, is it possible to install a specific instance of the RS in the working area of a particular machine, are the parameters of the connecting part RI parameters of the machine seat.

[00087] Compatibility rules must belong to certain contexts, i.e. be context sensitive. So the compatibility rules presented in the diagram should be activated only in the context (process) - "operation / installation of radiation in the machine chuck".

[00088] Ultimately, a semantic network is formed that defines the interconnections and compatibility of various objects included in various classification groups.

[00089] The compatibility of two objects in such a network is determined by the compatibility of their constituent parts belonging to the same contexts (processes).

[00090] Assume that all the necessary information about the classifications, attributes, contexts and relationships of objects is generated.

[00091] In order to perform a semantic search for objects in a semantic data structure, it is necessary to perform the following steps in the specified order:

[00092] Get the category of the desired object, the process in which the desired object and related objects involved in this process participate from the information input device, which may be a personal computer, tablet, not limited to.

[00093] Select all objects in the category of the desired object, at least one component of which is associated with the above process;

[00094] Among the objects selected in the previous step, select those objects whose components are compatible with the components of adjacent objects based on the compatibility rules applicable in the context of the above process. [00095] It is also possible to modify the described method:

[00096] in a first step, parameters of the desired object are further obtained, including at least one attribute;

[00097] after performing the basic steps, additionally select those objects that satisfy the parameters of the desired object.

[00098] According to the described methods, the search can be carried out by the data structure generated by the method described previously.

[00099] This method can be implemented in a hardware-software complex described below.

[000100] Using this invention in production can significantly improve the accuracy of the search for technological objects in an MDM system using a semantic data structure.

[000101] A person skilled in the art can readily understand other embodiments of the invention from the foregoing description. This application is intended to cover any use or application of the following general principles of the invention, and including those deviations from the present invention that appear within the scope of known or ordinary practice in the art. It is assumed that the description is considered only as an example, with the essence and scope of the present invention indicated by the claims.

[000102] It should be appreciated that the present invention is not limited to the precise structures that have been described above and illustrated in the accompanying drawings, and that various modifications and changes can be made without departing from its scope. It is intended that the scope of the invention be limited only by the appended claims.

Claims

CLAIM

1. The method of semantic search in the object-process data model includes the following steps:

• formulate the classification of objects and processes by creating ontologies, at the same time, in each category of objects, at least one component is distinguished from which the object consists, including at least the identifier of the component and one or more attributes characterizing this component;

• form a relationship between each component of the object with at least one process in which this component is involved;

• for each process find all pairs of objects that are associated with this process;

• form a relationship that defines the rules of interaction for the objects found in the previous step for each process;

• receive from the input device the category of the desired object and at least one process in which the desired object is involved;

• receive all objects that are associated with the above process.

2. The method according to claim 1, characterized in that when forming the classification of objects and processes by creating ontologies, where at least one component is distinguished in each category of objects from which the object consists, including at least the identifier of the component and one or more attributes characterizing a given component, the identifier is an integer value.

3. The method according to claim 1, characterized in that when forming a connection between each component of the object with at least one process in which this component is involved, the connection is set in the form of mathematical and logical operations.

4. The method according to claim 1, characterized in that when forming a connection between each component of the object with at least one process in which this component is involved, the component of the object is associated with one process.

5. The method according to claim 1, characterized in that when forming a connection between each component of the object with at least one process in which this component is involved, the component of the object is associated with several processes.

6. Method no n.l, characterized in that when forming a relationship that defines the rules of interaction for the found objects for each process, the object can have several internal states - structures, each of which can be associated with one or more processes