Knowledge Graph Construction Method for 3D Digital Model Design Data of Mechanical Products

Abstract

This invention discloses a method for constructing a knowledge graph from 3D digital model design data for mechanical products. It primarily addresses the problem that existing technologies cannot acquire, represent, and reuse 3D model design data. The implementation scheme is as follows: 1) Integrating and processing the 3D model data; 2) Constructing a knowledge acquisition program to extract a total knowledge set from the 3D model; 3) Representing and constraining the total knowledge set to obtain a processed knowledge set; 4) Instantiating and tree-forming the processed knowledge set to generate feature trees representing functions, structures, and constraints; 5) Analyzing and processing the obtained feature trees to form sets of structure-structure, structure-function, and structure-constraint relationships; 6) Importing the obtained feature trees and relationship sets into triplet form into Neo4j to generate a knowledge graph. This invention can extract, represent, and reuse design data from 3D digital models and can be used for intelligent design of structural schemes for mechanical products.

Description

Technical Field

[0001] This invention belongs to the field of intelligent design technology for mechanical products, and further relates to a method for constructing a knowledge graph to extract, represent and reuse design data from a three-dimensional digital model, which can be used for intelligent design of structural schemes for mechanical products. Background Technology

[0002] Mechanical product design companies accumulate a wealth of design experience data during design and production. This data often exists in unstructured form, including unstructured textual resources such as technical principles, design guidelines, and industry standards, as well as unstructured 3D digital model data. While methods for processing textual data are relatively mature, enabling knowledge reuse and unified management through knowledge representation, extraction, and construction, unstructured 3D digital model data is far more complex. It involves more than just simple geometric figures; it contains rich implicit design knowledge. This includes the product's components, the relationships between them, and model parameters. However, currently, due to a lack of effective methods for representing design knowledge, mechanical engineers typically rely on personal experience and extracting design concepts, common characteristics, and design knowledge from paper drawings, CAD drawings, or other design data to aid the product design process. In the absence of appropriate methods for representing and extracting model design knowledge, the unstructured 3D digital model data accumulated by companies during the design and production of structural components often fails to fully utilize its inherent design knowledge, and its potential value is frequently overlooked. Fully leveraging the value of this type of data to achieve knowledge representation and extraction from 3D models is of great significance for the high-quality construction of knowledge bases in the design field and the efficient reuse of design knowledge.

[0003] Harbin Institute of Technology disclosed a "Knowledge Representation Method and Driving System Based on Power Grid Artificial Intelligence Simulation Analysis" in its patent application (application number 202211230433.7). Firstly, addressing the diversity of human experience knowledge, it designed an SPO knowledge representation model to model the adjustment process, operations, and program calls. While unifying knowledge representation, it combines qualitative-quantitative and relational-logical knowledge to solve the problem of difficulty in constructing a knowledge-driven model due to the difficulty in knowledge representation. This invention realizes the intelligentization of the power grid digital simulation analysis process through a modular knowledge-driven framework, enabling the entire process of analysis and adjustment to be automated without human intervention. It extracts knowledge from a knowledge base using a knowledge extraction algorithm and performs knowledge cleaning, then uses the Neo4j graph database to visualize the constructed power grid digital simulation analysis knowledge graph. The effectiveness and feasibility of the knowledge-driven model were verified on an improved node system. However, due to the specific nature of its knowledge source, this method suffers from insufficient scalability of the knowledge representation and extraction model; it can only represent and extract processes, operations, and program calls within relevant industries, while struggling to fully represent and acquire the complex design knowledge contained in three-dimensional digital model data.

[0004] The Information Engineering University of the Strategic Support Force of the Chinese People's Liberation Army disclosed a "map knowledge representation method based on knowledge graphs" in its patent application (application number 202210770046.6). This method first acquires map data and extracts map knowledge from it, including conceptual knowledge and rule-based knowledge. The conceptual knowledge refers to common knowledge in the field of cartography and includes map element knowledge; the rule-based knowledge consists of various models, rules, and principles formed during mapmaking. Then, based on the map element knowledge, a map knowledge network graph is generated using a knowledge graph. Finally, knowledge reasoning is performed based on the rule-based knowledge to complete the map knowledge network graph, generating new entities and / or new entity relationships. The limitation of this method is that it only provides structured representation and construction for conceptual and rule-based knowledge in map data, without establishing corresponding knowledge extraction models for unstructured data. This makes it difficult to apply to the extraction of unstructured 3D digital model design data with complex characteristics. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of the existing technologies and research by proposing a knowledge graph construction method for three-dimensional digital model design data of mechanical products. This method acquires, represents, and processes the three-dimensional digital model design data of mechanical products, improves the reusability of design data, and promotes the development of intelligent design of mechanical products empowered by design data.

[0006] To achieve the above objectives, the technical solution of the present invention includes the following steps:

[0007] (1) Integration and processing of 3D digital model data resources:

[0008] (1a) Integrate and organize the three-dimensional digital model data of similar structural components and use them as a knowledge source for unstructured data;

[0009] (1b) The modeling process and feature operation naming are standardized sequentially for the three-dimensional digital model data in the knowledge source to obtain the processed data set;

[0010] (2) Obtain the total knowledge set:

[0011] (2a) Open the 3D model in the preprocessed dataset in existing modeling software;

[0012] (2b) Build an automated knowledge acquisition program in the programming software to acquire knowledge from the 3D model data that has been opened in the modeling software, and obtain the total knowledge set TM hidden in the 3D model, which includes information such as part name, part geometry, sketch, constraints and features;

[0013] (3) Represent and constrain the knowledge set TM:

[0014] (3a) The elements in the knowledge set TM are expressed and organized using a joint design mechanism model of function-structure-constraint, resulting in the represented knowledge set TM′;

[0015] (3b) Constrain the knowledge elements in the knowledge set TM′ to obtain the constrained knowledge set TM″;

[0016] (4) Construct the functional feature tree, structural feature tree, and constraint feature tree:

[0017] (4a) Extract the structural element SE and constraint element CE from the knowledge set TM″;

[0018] (4b) Integrate all structural elements SE and constraint elements CE to obtain the structural element set SM and the constraint element set CM;

[0019] (4c) Add a function tag to each structure element SE in the structure element set SM to form a function element set FM corresponding to the structure element set SM;

[0020] (4d) Combining the hierarchical information represented on the modeling tree of the 3D model itself, the functional element set FM, structural element set SM and constraint element set CM are traversed respectively. Then, each element node after traversal is instantiated into an object container OD so that the node has the characteristic of representing the feature sequence information of the 3D digital model. After reintegration, the instantiated FM′, SM′ and CM′ are obtained.

[0021] (4e) Using the treelib standardization module, the instantiated FM′, SM′ and CM′ data are transformed into a tree structure that is easy to query and expand nodes, resulting in the functional feature tree FT, structural feature tree ST and constraint feature tree CT.

[0022] (5) Construct the set of structure-structure relationships, the set of structure-function relationships, and the set of structure-constraint relationships:

[0023] (5a) Expand the structural feature tree ST with the relInfo field as the relation value, and after integration, obtain the structure-structure relation set S-SM;

[0024] (5b) Expand the functional feature tree FT with the relInfo field as the relation value, and then match the structural elements in the structure tree ST with the functional elements one by one to obtain the structure-function relation set S-FM;

[0025] (5c) Expand the constraint feature tree CT with the relInfo field as the relation value, and then match the structural elements in the structure tree ST with the constraint elements one by one to obtain the structure-constraint relation set S-CM;

[0026] (6) Construct a domain knowledge graph for 3D digital model design data:

[0027] (6a) The data obtained above in the structural feature tree ST, functional feature tree FT, constraint feature tree CT, structure-structure relation set S-SM, structure-function relation set S-FM, and structure-constraint relation set S-CM are uniformly integrated into the triple representation of {head entity, relation, tail entity}.

[0028] (6b) Import the data in triplet representation into existing knowledge graph construction software to obtain a domain knowledge graph of the three-dimensional digital model design data.

[0029] Compared with the prior art, the present invention has the following advantages:

[0030] First, the design ensures high efficiency and quality in knowledge extraction.

[0031] The proposed design data acquisition method based on joint representation of function, structure, and constraint decomposes the hidden data of the 3D digital model of structural components into multiple levels and represents it in the form of a tree structure, until the structural feature information is decomposed into the finest-grained leaf node information, such as holes and plates. This refined structural decomposition and representation enables higher efficiency and quality in the representation and extraction of 3D digital model design data.

[0032] Second, it enables efficient management, reuse, and expansion of 3D digital model design data.

[0033] This invention proposes a method that combines function, structure, and constraint to parse the 3D model of a structural component into a structural feature tree. By using a tree structure to organize the functional and constraint information of the structural component, modularization of structural component design knowledge is achieved, and this knowledge is constructed into a knowledge graph form, enabling efficient management, reuse, and expansion of design knowledge. Attached Figure Description

[0034] Figure 1 This is a flowchart illustrating the implementation of the present invention;

[0035] Figure 2 This is a three-dimensional model diagram of an embodiment of the present invention;

[0036] Figure 3 This is a comparison diagram of the geometric properties of the parts before and after preprocessing in an embodiment of the present invention;

[0037] Figure 4 This is a diagram of the secondary development interface of the modeling software used in the embodiments of the present invention;

[0038] Figure 5 This is a diagram illustrating the functional-structural-constraint joint design mechanism of this invention.

[0039] Figure 6 This is an example diagram of the feature tree representing the structure, function, and constraints in an embodiment of the present invention;

[0040] Figure 7 This is an example diagram of a knowledge graph in an embodiment of the present invention;

[0041] Figure 8 This is an example diagram of structure-function mapping query in an embodiment of the present invention. Detailed Implementation

[0042] The embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0043] Reference Figure 1 The implementation steps of this method are as follows:

[0044] Step 1: Integrate and process the 3D digital model data resources.

[0045] 1.1) Integrate and organize the 3D digital model data of similar structural components, and use them as a source of knowledge for unstructured data;

[0046] In this embodiment of the invention, a three-dimensional digital model of a type of mechanical product from a certain research institute is used as a source of unstructured data knowledge.

[0047] This 3D digital model dataset contains 3D digital models of four similar structural components. This invention uses one of these structural component models for example verification of the solution. The 3D model is as follows: Figure 2 As shown, its model information is shown in Table 1.

[0048] Table 1. Overview of Model Information

[0049]

[0050] 1.2) The 3D numerical model data from the knowledge source is subjected to modeling process standardization and feature operation naming standardization in sequence to obtain the preprocessed dataset:

[0051] Different designers exhibit varying modeling approaches and methods when modeling similar structural components. These differences primarily manifest in the modeling process itself, rather than in functional implementation. To establish a universal knowledge representation method and resolve modeling discrepancies among similar structural components, it is necessary to standardize the modeling process for such components. Furthermore, in molded model design, the naming of feature operation attributes is often based on the functional names of the modeling software itself, such as hole commands, fillets, and bosses. Considering the need for knowledge reuse, these names themselves lack practical meaning; therefore, it is necessary to correlate feature operation names with the actual structural names of the components. For example, the hole operation can be mapped to a specific structural "support hole," thus imbuing it with practical meaning for subsequent knowledge reuse and inference.

[0052] Based on this, this example standardizes the modeling process and feature operation naming for 3D digital models, and compares the geometric attributes of the parts before and after the processing, such as... Figure 3 As shown.

[0053] Step 2: Obtain the total knowledge set.

[0054] 2.1) Query relevant technical documents to assist in the extraction of the overall knowledge set:

[0055] 2.1.1) Consult the secondary development documentation of the existing modeling software CATIA, such as... Figure 4 As shown in Table 2, some commonly used API interfaces are listed below.

[0056] Table 2 Commonly Used APIs for Secondary Development of Modeling Software

[0057]

[0058] 2.1.2) Consult the technical documentation on the interaction between the programming software Python and the modeling software CATIA. This includes the fact that the win32.com module is commonly used in the Windows environment to complete the interaction between the programming software and the modeling software, and that in the Windows environment, most applications, operating system internals, and third-party libraries widely use the win32.com module to realize data interaction between the programming software and the modeling software.

[0059] 2.2) Open the 3D model in the processed dataset in the existing modeling software CATIA;

[0060] 2.3) Construct an automated knowledge acquisition program to complete the acquisition of the total knowledge set:

[0061] 2.3.1) An interactive environment between the programming software Python and the modeling software CATIA is built using the win32com module;

[0062] 2.3.2) Using the software secondary development interface documentation provided by the modeling software CATIA, in the interactive environment built above, the interface for information such as part name, part geometry, parameters, sketch, constraints, features, and materials is recursively called through the host object to obtain the total set of design knowledge TM, as shown in Table 3.

[0063] Table 3: Overview of the Knowledge Collection

[0064]

[0065] Step 3: Represent and constrain the knowledge set TM.

[0066] 3.1) The elements in the knowledge set TM are expressed and organized using a joint design mechanism model of function, structure, and constraint, resulting in the represented knowledge set TM′:

[0067] The function-structure-constraint joint design mechanism model is an abstract generalization of the principles of part structural design. It divides knowledge into three parts: structural elements, constraint elements, and functional elements, and ultimately represents this knowledge using a tree structure, such as... Figure 5 As shown. The structural feature tree represents the process of analyzing the structural elements of a specific part; the functional feature tree represents the functional elements corresponding to each structural element in the structural feature tree of a part. The overall function is analyzed layer by layer in a tree format into individual functional elements, which correspond to the structural elements. That is, the functional feature tree corresponds to the structural feature tree, achieving a mutual mapping between structure and function. The function implemented by each part is manifested in a specific environment; therefore, the actual assembly environment of the part needs to be considered, and the assembly constraints of the part need to be analyzed. Each structural element has a corresponding constraint, called a constraint element. The process of organizing constraint elements is also the process of forming the constraint feature tree.

[0068] Based on the above principles, the knowledge set TM extracted in step 2.3.2) is divided into structural, constraint, and functional elements. The design knowledge in TM is reorganized in the form of function-structure-constraint to obtain the represented knowledge set TM′, as shown in Table 4.

[0069] Table 4 shows a summary of the knowledge set.

[0070]

[0071]

[0072] In the table above, Null indicates that the current attribute value is temporarily empty.

[0073] 3.2) Use the ontology to constrain the knowledge elements in the knowledge set TM′ to obtain the constrained knowledge set TM″;

[0074] The ontology is a display description of a shared conceptual model. It is an abstract generalization of the essence of an entity's existence from a specific domain. It is a set of concepts that contains the basic elements and relationships between elements within a certain domain. It is an abstraction of the operating mechanism of that domain and has universality.

[0075] This example applies ontology to plan the knowledge ontology of 3D digital model design. There are 5 types of ontology, including ontology at the levels of structural components, structural elements, constraint elements, functional elements, and material elements, as shown in Table 5.

[0076] Table 5 Types of Design Knowledge Ontologies

[0077]

[0078] The five types of ontology are used to constrain the structural element part SP, the constraint element part CP, the functional element part FP, and the other element part OP in the knowledge set TM′, respectively, to obtain the constrained SP′, CP′, FP′, and OP′. After reintegration, the constrained knowledge set TM″ is obtained, as shown in Table 6.

[0079] Table 6. Overview of the Constrained Knowledge Set

[0080]

[0081] In the table above, Null indicates that the current attribute value is temporarily empty.

[0082] Step 4: Construct the functional feature tree, structural feature tree, and constraint feature tree.

[0083] 4.1) Extract the structural element SE and constraint element CE from the constrained knowledge set TM″, and integrate all structural elements SE and constraint elements CE to obtain the structural element set SM and the constraint element set CM:

[0084] SM = {clamp plate, clamp hole, clamp chamfer, ...}

[0085] CM = {Support surface, clamp cable, reference, ...};

[0086] 4.2) Add a function tag to each structural element SE in the structural element set SM, such as the function of the clamp hole is to fix the clamp plate, forming a function element set FM corresponding to the structural element set SM:

[0087] FM = {Cable support, clamp plate fixation, clamp plate stress improvement, ...};

[0088] 4.3) Combining the hierarchical information represented on the modeling tree of the 3D model in this embodiment, the functional element set FM, structural element set SM, and constraint element set CM are traversed respectively. Then, each element node after traversal is instantiated into an object container OD. Some results are shown in Table 7:

[0089] Table 7 shows the results of node instantiation.

[0090]

[0091] In the table above, the unique ID of the node is automatically generated using the uuid module, and null indicates that the content of the segment is empty after instantiation.

[0092] 4.4) After reorganizing the instantiated object container nodes, we get the instantiated FM′, SM′, and CM′;

[0093] 4.5) Using the treelib module, the aggregated data of FM′, SM′, and CM′ are transformed into tree structures that are easy to represent and convenient for fast searching, resulting in the functional feature tree (FT), structural feature tree (ST), and constraint feature tree (CT). The transformed results are as follows: Figure 6 As shown.

[0094] Step 5: Construct a set of structure-structure relationships, structure-function relationships, and structure-constraint relationships.

[0095] 5.1) Expand the structural feature tree ST using the relInfo field as the relation value, and after integration, obtain the structure-structure relation set S-SM, as shown in Table 8.

[0096] Table 8 shows a partial display of the structure-structure relationship set.

[0097]

[0098]

[0099] 5.2) Expand the functional feature tree FT with the relInfo field as the relation value, and then match the structural elements in the structural feature tree ST with the functional elements one by one to obtain the structure-function relation set S-FM, as shown in Table 9.

[0100] Table 9 shows the structure-function relationship set.

[0101]

[0102] 5.3) Expand the constraint feature tree CT with the relInfo field as the relation value, and match the structural elements in the structural feature tree ST with the constraint elements one by one to obtain the structure-constraint relation set S-CM, as shown in Table 10.

[0103] Table 10 shows a partial display of the structure-constraint relationship set.

[0104]

[0105] Step 6: Construct a knowledge graph for the design of 3D digital models.

[0106] 6.1) The data obtained above in the structural feature tree ST, functional feature tree FT, constraint feature tree CT, structure-structure relation set S-SM, structure-function relation set S-FM, and structure-constraint relation set S-CM are uniformly integrated into the {head entity, relation, tail entity} triple representation, with a total of 23 triple pairs. Table 11 shows the representation of 8 of these pairs.

[0107] Table 11 shows the partial triplet set in this embodiment.

[0108]

[0109] 6.2) Import the data in triplet representation into Neo4j, an existing software for constructing knowledge graphs, to obtain the knowledge graph of the 3D digital model design data in this embodiment, such as... Figure 7 As shown.

[0110] The effectiveness of this invention can be further illustrated by the verification results of the spectra:

[0111] Suppose we want to design a structure that satisfies the function of "connecting support surfaces". The corresponding Cypher statement is: "MATCH(r)-(:structure function)->(n:function element)RETURN r", and the query result is as follows. Figure 8 In this context, each circle represents an entity, and arrows between entities indicate a relationship between them.

[0112] from Figure 8 As can be seen, the support plate can connect to the support surface, thus enabling the reuse of the 3D digital model design data support plate. The results of this atlas verification show that the present invention can achieve efficient retrieval and reuse of design knowledge.

[0113] The above description is merely a specific example of the present invention and does not constitute any limitation on the present invention. The reference numerals for each step are only for clear description of the technical solution of the present invention, and their order is not limited. Obviously, those skilled in the art, after understanding the content and principles of the present invention, can make various modifications and changes in form and detail without departing from the principles and structure of the present invention. However, these modifications and changes based on the concept of the present invention are still within the scope of protection of the claims of the present invention.

Claims

1. A method for constructing a knowledge graph for three-dimensional digital model design data of mechanical products, characterized in that, Including the following: (1) Integrating and processing 3D digital model data resources: (1a) Integrate and organize the three-dimensional digital model data of similar structural components and use them as a knowledge source for unstructured data; (1b) The modeling process and feature operation naming are standardized sequentially for the three-dimensional digital model data in the knowledge source to obtain the processed data set; (2) Obtain the total knowledge set: (2a) Open the 3D model in the preprocessed dataset in existing modeling software; (2b) Build an automated knowledge acquisition program in programming software to acquire knowledge from the 3D model data already opened in the modeling software, and obtain the total set of knowledge hidden in the 3D model. This includes information such as part name, part geometry, sketch, constraints, and features; (3) The total set of knowledge Knowledge representation and constraints: (3a) Employing a combined function-structure-constraint design mechanism model for the total knowledge set The elements in the data are expressed and organized to obtain the total represented knowledge set. ; (3b) On the total set of knowledge By constraining the knowledge elements in the dataset, we obtain the constrained total knowledge set. ; (4) Construct the functional feature tree, structural feature tree, and constraint feature tree: (4a) From the total collection of knowledge Filter and extract structural elements and constraint elements ; (4b) Integrate all structural elements Constraint elements , to obtain the set of structural elements and constraint element set ; (4c) For the set of structuring elements Each structural element Add function tags one by one to form a set of structural elements. Functional element set ; (4d) Combining the hierarchical information represented on the modeling tree of the 3D model itself, the functional element sets are respectively... , set of structural elements and constraint element set Perform the traversal, and then instantiate each element node as an object container. This allows nodes to possess the characteristic of representing the feature sequence information of a three-dimensional digital model, and after reintegration, an instantiated version is obtained. , as well as The object container Its descriptive formula is as follows: ； in, Indicates the unique identifier of the node. Indicates the node type. Indicates current node information, This represents the child node information of the current node. This indicates the relationship information of the current node; (4e) Utilization Standardized modules will be instantiated , and The data is transformed into a tree structure that is easy to query and expand nodes, resulting in a functional feature tree. Structural feature tree and constrained feature tree ; (5) Construct the set of structure-structure relationships, the set of structure-function relationships, and the set of structure-constraint relationships: (5a) Structural feature tree by Fields are expanded as relation values ​​and integrated to obtain a set of structure-structure relations. ; (5b) Functional feature tree by Expand the fields as relational values, then construct the structure tree. By mapping each structural element to a functional element, a structure-function relationship set can be obtained. ; (5c) Constraint feature tree by Expand the fields as relational values, then construct the structure tree. The structural elements and constraint elements are mapped one by one to obtain the structure-constraint relationship set. ; (6) Construct a domain knowledge graph for 3D digital model design data: (6a) Obtain the structural feature tree as described above Functional feature tree Constraint feature tree Structure-structure relationship set Structure-Function Relationship Set Structure-constraint relationship set The data in the data is uniformly integrated into a triple representation of {head entity, relation, tail entity}; (6b) Import the data in triplet representation into existing knowledge graph construction software to obtain a domain knowledge graph of the three-dimensional digital model design data.

2. The method according to claim 1, characterized in that, Step (1b) standardizes the modeling process and the naming of feature operations in the knowledge source data. The steps include the following: (1b1) Open the 3D model to be preprocessed in the modeling software; (1b2) Adjust the order of modeling feature operations of the 3D model in the modeling software to keep the modeling order of the same type of 3D model consistent, and complete the standardization of its modeling process; (1b3) Modify the attribute names of modeling feature operations in the modeling software to give the feature operations actual engineering representation meaning and complete the standardization of their feature operation names.

3. The method according to claim 1, characterized in that, Step (2b) involves constructing an automated knowledge acquisition program to obtain the total knowledge set. The implementation steps include the following: (2b1) Through The module establishes an interactive environment between programming and modeling software. (2b2) ​​Using the software secondary development interface documentation provided by the modeling software, in the environment built above, the interface for information such as part name, part geometry, parameters, sketch, constraints, and features is recursively called through the host object to obtain the total set of design knowledge. : ； in Indicated as product name, This is represented as product parameter information. Represented as a sketch group, , , These are respectively represented as entity feature set, boundary feature set, and mixed feature set.

4. The method according to claim 1, characterized in that, In step (3a), a joint design mechanism model of function, structure, and constraint is used to... The elements in the text are expressed and organized to form the total knowledge set extracted in step (2b). Divide the structure, constraints, and functional elements, and reorganize them in a structure-constraint-function format. The design knowledge in the data is used to obtain the total set of knowledge after representation. : ； in , , , These are respectively represented as the structural element part, constraint element part, functional element part, and other element parts after organization.

5. The method according to claim 1, characterized in that, In step (3b), the total knowledge set The knowledge elements in the data are constrained, including five types of ontological constraints. These five types of ontological constraints include constraints at the levels of structural components, structural elements, constraint elements, functional elements, and material elements. The implementation steps include the following: (3b1) For the total set of knowledge respectively The structural element part Constraint element part Functional elements Other elements Apply constraints to obtain the constrained result. , , , ; (3b2) Integrating structural elements Constraint elements Functional elements Other elements The total set of knowledge after obtaining constraints ; ； in , , , These are respectively represented as the constrained structural element part, the constraint element part, the functional element part, and the other element parts.

6. The method according to claim 1, characterized in that, In step (4b), all structuring elements SE and constraint elements CE are integrated to obtain the structuring element set SM and the constraint element set CM, which are represented as follows: ； in, Represents the first in the set of structural elements One structural element, Represents the first constraint element in the set. Constraint elements, The range is {1, }, This indicates the size of the set of feature elements obtained in step (2b).

7. The method according to claim 1, characterized in that, In step (4d), the instantiated version is obtained. , as well as Its descriptive formula is as follows: ； in, , , They represent the instantiated first and second generations respectively. Each object container element consists of a functional object container element, a structural object container element, and a constraint object container element. The range is {1, }, This indicates the size of the set of feature elements obtained in step (2b).

Images