A lightweight self-describing packaging method based on long-term preservation of BIM models

Abstract

The application provides a lightweight self-description packaging method based on long-term preservation of a BIM model, and relates to the technical field of lightweight BIM models. Through S1: constructing an electromechanical equipment engineering value evaluation model, the equipment components in the BIM model are quantitatively evaluated based on three dimensions; S2: according to the equipment comprehensive value score, the equipment components are mapped to a preset LOD archiving level, and the BIM model is subjected to semantic-aware lightweight processing; S3: the data after the lightweight processing is organized according to a ternary separation structure; S4: an archive information package is constructed, and the geometric resource library, the attribute resource library and the semantic relationship library are packaged together with the archive information package to generate a self-description archive package; the problem that traditional lightweight processing only focuses on geometric complexity is solved, on-demand lightweight processing is realized, the lightweight degree is automatically adjusted according to the equipment value, the semantic information integrity of the key equipment is improved, and the long-term readability of the information is improved.

Description

Technical Field

[0001] This invention relates to the technical field of lightweight BIM models, and in particular to a lightweight self-describing encapsulation method based on long-term storage of BIM models. Background Technology

[0002] With the widespread application of Building Information Modeling (BIM) in the field of mechanical and electrical engineering, the demand for the long-term preservation and efficient utilization of BIM models of mechanical and electrical equipment, as important digital archives for project completion, operation and maintenance management, and audit traceability, is becoming increasingly urgent.

[0003] Existing BIM model archiving and storage technologies have significant drawbacks: First, while relying on IFC open standards for model data exchange supports multi-party collaborative access management and data review, they fail to lightweight the model, resulting in massive amounts of original geometric and attribute data. This leads to high storage consumption, low transmission and retrieval efficiency, and fails to resolve the cost-availability conflict for long-term preservation of large-scale electromechanical equipment models. Second, geometry export schemes based on general 3D formats export BIM model geometry information to lightweight formats such as glTF and OBJ, reducing data volume and improving visualization efficiency to some extent. However, they only retain geometric meshes and basic appearance information, significantly losing non-geometric business information such as engineering semantic attributes, technical parameters, and system relationships specific to electromechanical equipment. Third, commercial software-built-in archiving solutions heavily rely on specific vendor platforms, violating the principle of technology neutrality and failing to guarantee long-term data readability. Summary of the Invention

[0004] The main objective of this invention is to provide a lightweight self-describing encapsulation method for long-term storage of BIM models, which solves the three major contradictions between full storage and availability, lightweight design and semantic integrity, and vendor binding and long-term storage.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a lightweight self-describing encapsulation method based on long-term storage of BIM models, comprising the following steps: S1: Construct a value assessment model for electromechanical equipment engineering, and quantitatively assess the equipment components in the BIM model based on three dimensions: equipment importance, equipment service life, and semantic demand intensity, and generate a comprehensive equipment value score; S2: Based on the comprehensive value score of the equipment, map the equipment components to the preset LOD archiving level, and perform semantically aware lightweight processing on the BIM model according to the lightweight rule library corresponding to each level to generate a lightweight geometric model and a structured attribute dataset. S3: Organize the lightweight data according to a ternary separation structure, and store geometric information, attribute information and semantic relationship information in open standard formats to form a geometric resource library, an attribute resource library and a semantic relationship library; S4: Construct an archive information package, record the encapsulation specification version, file path, project metadata and file role description, and encapsulate the geometric resource library, attribute resource library, semantic relation library and archive information package together to generate a self-describing archive package.

[0006] In the preferred embodiment, the comprehensive value scoring function of the electromechanical equipment engineering value assessment model is: V = α·I + β·L + γ·S; Where V is the overall value score of the equipment; I is the importance of the equipment; L is the equipment usage cycle; S is the semantic demand intensity; and α, β, and γ are preset weight coefficients.

[0007] In the preferred embodiment, the equipment importance, equipment usage cycle, and semantic demand intensity are each calculated by weighting multiple sub-indicators, including: The importance of equipment is determined by a weighted average of its impact on system functionality, maintenance difficulty coefficient, and economic value. The equipment service life is obtained by weighting the design life, expected service life, and replacement frequency. The strength of semantic requirements is obtained by weighting the dependence of business systems, the frequency of semantic data usage, and the requirements for information integrity.

[0008] In the preferred embodiment, the LOD archiving level corresponds to different geometric precision and attribute retention granularity, including: LOD Archive 500 preserves the highest precision geometry and all attributes; LOD Archive 400 preserves the main geometric features and key business attributes; LOD Archive 300 retains the basic shape and logo information; LOD Archive 200 preserves system connections and key parameters, and geometrically simplifies them into symbolic representations.

[0009] In a preferred embodiment, the ternary separation structure includes: The geometric resource library uses the glTF 2.0 format to store lightweight 3D geometric information; The attribute resource library uses JSON-LD format to store non-geometric attribute information and supports semantic tags and links to external knowledge bases; A semantic relational database uses RDF format to store the relationships between components and constructs a project knowledge graph.

[0010] In the preferred embodiment, the processing of the geometric resource library includes: adaptive patch simplification based on LOD level, material texture compression, vertex data optimization, similar volume merging, and node hierarchy preservation.

[0011] In the preferred embodiment, the processing of the attribute resource library includes: structured attribute extraction, key attribute filtering, semantic tag annotation, and linking to external knowledge bases.

[0012] In the preferred embodiment, the processing of the semantic relation library includes: component association identification, RDF triple generation, relation ontology definition, and knowledge graph construction.

[0013] In a preferred embodiment, the file information package includes at least: Unique identifier and creation time of the file package; Packaging specification version number; The path, format, and role description of each resource repository file; Metadata such as project name, archiving unit, and archiving date.

[0014] In a preferred embodiment, the method further includes: parsing the JSON-LD and RDF files in the archive package to query and reason about device attributes and their semantic relationships, supporting intelligent retrieval and knowledge graph construction; The self-describing archive is a standard ZIP archive and is identified using a specific file extension.

[0015] This invention provides a lightweight self-describing encapsulation method for long-term storage of BIM models. The method involves: S1: Constructing a value assessment model for electromechanical equipment engineering, quantitatively evaluating equipment components in the BIM model based on three dimensions, and generating a comprehensive equipment value score; S2: Mapping equipment components to a preset LOD (Level of Detail) archiving level based on the comprehensive equipment value score, and performing semantically aware lightweight processing on the BIM model; S3: Organizing the lightweight processed data according to a three-element separation structure to form a geometric resource library, an attribute resource library, and a semantic relation library; S4: Constructing an archive information package and encapsulating the geometric resource library, attribute resource library, semantic relation library, and archive information package together to generate a self-describing archive package. This method overcomes the problem of traditional lightweighting focusing only on geometric complexity, achieving on-demand lightweighting, automatically adjusting the degree of lightweighting according to equipment value, improving the semantic information integrity of key equipment, and enhancing the long-term readability of information. Attached Figure Description

[0016] The present invention will be further described below with reference to the accompanying drawings and embodiments: Figure 1 This is a flowchart of the lightweight self-describing encapsulation method of the present invention; Figure 2 This is a schematic diagram of the overall process of the lightweight self-describing encapsulation method of the present invention. Detailed Implementation

[0017] Example 1 like Figure 1-2As shown, a lightweight self-describing encapsulation method based on long-term storage of BIM models includes the following steps: S1: Construct a value assessment model for electromechanical equipment engineering, and quantitatively assess the equipment components in the BIM model based on three dimensions: equipment importance, equipment usage cycle, and semantic demand intensity, to generate a comprehensive equipment value score.

[0018] S2: Based on the comprehensive value score of the equipment, map the equipment components to the preset LOD archiving level, and perform semantically aware lightweight processing on the BIM model according to the lightweight rule library corresponding to each level, generating a lightweight geometric model and a structured attribute dataset.

[0019] S3: Organize the lightweight data according to a ternary separation structure, and store geometric information, attribute information and semantic relationship information in open standard formats to form a geometric resource library, an attribute resource library and a semantic relationship library.

[0020] S4: Construct an archive information package, recording the encapsulation specification version, file path, project metadata, and file role description. Encapsulate the geometric resource library, attribute resource library, semantic relation library, and archive information package together to generate a self-describing archive package.

[0021] This embodiment quantifies equipment value by constructing a value assessment model, maps LOD levels and lightweights them. Through value assessment-driven semantic awareness lightweighting, data volume can be significantly reduced while preserving key semantics, lowering storage and transmission costs. Based on the W3C open standards (glTF, JSON-LD, RDF), a three-element separation structure avoids dependence on specific BIM software vendors. Engineering semantics (equipment parameters, system associations, business relationships) are stored separately from geometry, upgrading BIM archives into computable knowledge assets, enabling full lifecycle operation and maintenance. The self-describing archive package, using JSON-LD and RDF formats, allows BIM data to be integrated into the semantic web, improving compatibility and data interoperability.

[0022] This embodiment uses the BIM model archiving process of the "main transformer" of a pumped storage power station as an example to explain in detail the implementation steps of the present invention.

[0023] Step S1: Semantic awareness layer, device value assessment and lightweight decision-making.

[0024] In the preferred solution, the three dimensions—equipment importance, equipment usage cycle, and semantic demand intensity—are calculated by weighting multiple sub-indicators, including: The importance of equipment is determined by a weighted average of its impact on system functionality, maintenance difficulty coefficient, and economic value. The equipment service life is obtained by weighting the design life, expected service life, and replacement frequency. The strength of semantic requirements is obtained by weighting the dependence of business systems, the frequency of semantic data usage, and the requirements for information integrity.

[0025] The system loads the original BIM model of the main transformer (in IFC or .rvt format) and automatically identifies the equipment type as a main transformer. It then calls the electromechanical equipment engineering value assessment model to perform a quantitative assessment from three dimensions: S101: Calculation of Equipment Importance Index (I): I1 (Degree of impact on system function): The main transformer is the core equipment of the power station and is scored 10 points; I2 (Repair Difficulty Level): The repair is complex and requires a power outage. The score is 8 points. I3 (Economic Value): The equipment is expensive and has high procurement costs, with a score of 10. The calculation yields: I = 0.4 × 10 + 0.3 × 8 + 0.3 × 10 = 9.4 points.

[0026] S102: Calculation of Equipment Service Life Index (L): L1 (Equipment Design Life): Design life up to 30 years, rated 10 points.

[0027] L2 (Expected Lifespan): Expected lifespan is 30 years, with a score of 9.

[0028] L3 (Update Frequency): The main transformer has a low update frequency, and the score is 5 points.

[0029] The calculation yields: L = 0.4 × 10 + 0.3 × 9 + 0.3 × 5 = 8.2 points.

[0030] S103: Calculation of Equipment Service Life Index (L): S1 (Business System Dependency): Closely related to the operation and maintenance system and SCADA system, with a score of 10; S2 (Semantic Data Usage Frequency): Technical parameters are frequently queried, with a score of 9. S3 (Information Integrity Requirement): Operations and maintenance require complete parameters, manuals, and test reports, with a score of 8 points; The calculation yields: S = 0.4 × 10 + 0.3 × 9 + 0.3 × 8 = 9.1 points.

[0031] S103: Calculation of the overall value score (V).

[0032] In the preferred scheme, the comprehensive value scoring function of the electromechanical equipment engineering value assessment model is: V = α·I + β·L + γ·S; Where V is the overall value score of the equipment; I is the importance of the equipment; L is the equipment usage cycle; S is the semantic demand intensity; and α, β, and γ are preset weight coefficients.

[0033] Take the weight coefficient α=0.4, β=0.3, γ=0.3α=0.4, β=0.3, γ=0.3 The calculation yields: V = 0.4 × 9.4 + 0.3 × 8.2 + 0.3 × 9.1 = 8.95 points.

[0034] In this embodiment, the value of the equipment is quantified into a calculable mathematical expression, thereby achieving on-demand lightweighting, precise retention of core equipment, significant simplification of secondary equipment, and improved compression efficiency.

[0035] This embodiment introduces semantically aware lightweight rules based on LOD (Level of Detail) archiving levels. It predefines the data granularity to be retained at different archiving levels for electromechanical equipment based on its importance classification. According to the predetermined rule base, each component in the BIM model is automatically lightweighted, generating a corresponding lightweight geometric model and structured attribute dataset.

[0036] In the preferred scheme, the LOD archiving levels include: LOD Archive 500 (Complete Archive Level): Retains the highest precision geometry and all attributes. Suitable for core, expensive equipment requiring meticulous maintenance (such as main transformers, water pumps, and turbines).

[0037] LOD Archive 400 (Standard Archive Level): Retains key geometric features and critical business attributes. Applicable to most major equipment (such as distribution cabinets and air conditioning units).

[0038] LOD Archive 300 (Basic Archive Level): Retains basic shape and identification information. Suitable for secondary equipment, pipes, ducts, etc.

[0039] LOD Archive 200 (System Reference Level): Retains only system connectivity and key parameters, with geometric simplification to symbols. Used to represent system integrity and spatial routing.

[0040] In this embodiment, according to the preset LOD archive level mapping rules, devices with a comprehensive value score V≥8.5 correspond to LOD Archive 500 (complete archive level). Therefore, the system determines that the main transformer is suitable for LOD Archive 500 level and calls the corresponding lightweight rule library for processing.

[0041] The lightweighting rules used in this embodiment are shown in Table 1.

[0042] Table 1 Lightweighting Rules

[0043] Step S2: Lightweight data processing with ternary separation. The lightweight processed data is separated into three independent parts using open standards according to its internal logic, and then encapsulated together.

[0044] As shown in Table 2, this embodiment adopts a three-element separation data model, which splits the BIM archive into three logically independent parts that are stored using open standards.

[0045] Table 2 Data Model for Ternary Separation

[0046] In the preferred embodiment, the ternary separation structure includes: 1. Geometric resource library, using glTF 2.0 format, stores lightweight 3D geometric information; simplified triangle faces (dynamically adjusted based on semantic value), compressed materials and textures, optimized vertex data, and preserved node hierarchy structure.

[0047] Each electromechanical component (such as a circuit breaker) corresponds to a node in the geometric resource library and has a unique identifier.

[0048] 2. Attribute resource library, which uses JSON-LD format to store non-geometric attribute information and supports semantic tags and links to external knowledge bases; structured attribute storage, semantic tag definition, attribute relationship description, and links to external knowledge bases.

[0049] 3. Semantic relation library, which uses RDF format to store the relationships between components: triple relation storage, ontology definition, knowledge graph construction, semantic query capability, etc., to build the project knowledge graph.

[0050] The data processing flow for ternary separation is as follows.

[0051] In the preferred scheme, the processing of the geometric resource library includes: In accordance with LOD Archive 500, the system retains medium-precision geometric models of core components such as the transformer body, heat sink, bushing, oil conservator, and terminal box.

[0052] An adaptive simplification algorithm is used to preserve key geometric features (such as the outline of the heat sink) while simplifying microscopic details such as bolts, nameplate textures, and small-diameter round holes.

[0053] A similarity recognition algorithm is applied to optimize the geometric reuse of recurring sleeves of the same specification.

[0054] After processing, export the geometric data as a glTF 2.0 format file named transformer_01.gltf, and assign a globally unique identifier GUID: T-2024-001 to the transformer model nodes.

[0055] In the preferred solution, attribute data processing (generate an attribute repository): The system extracts all attribute fields from the original BIM model and retains all key attributes according to the LOD Archive 500 rules.

[0056] The attribute data is structured and organized in JSON-LD format and semantic tags are added for machine understanding and interconnection.

[0057] An example of the content of the generated JSON-LD file attributes_T-2024-001.jsonld is as follows: / / attributes_T-2024-001.jsonld {"@id": "urn:uuid:T-2024-001", "type": "transformer", "name": "Main Transformer - T1", "model": "SFP-980000 / 330", "rated capacity": {"value": "980000", "unit code": "KVA"}, "voltage ratio": "330 / 15.75 kV", "manufacturer": "TBEA", "installation date": "2024-08-10", "last inspection date": "2025-05-20"}.

[0058] In the preferred solution, semantic relationship processing (generate a semantic relationship repository): The system automatically analyzes relationships such as system associations, spatial connections, and document attachments in the model.

[0059] Identify and construct RDF triples to generate a Turtle format file relationships.ttl. An example of the content is as follows: The system automatically analyzes the model to generate an RDF triple file (relationships.ttl): <urn:uuid:T-2024-001><http: / / bimschema.org / partOf><urn:uuid:Electrical System>. <urn:uuid:T-2024-001><http: / / bimschema.org / isConnectedTo><urn:uuid:Busbar-330kV>. <urn:uuid:T-2024-001><http: / / bimschema.org / hasDocument><urn:uuid:TestReport-2024>.

[0060] In this embodiment, the ternary separation structure improves the computability of data.

[0061] Step S3: Generate a self-describing archive package.

[0062] In the preferred solution, the above three resource libraries, together with the archive information package, are packaged into a standard ZIP compression package, and it is recommended to use a specific file extension. The archive information package at least includes: 1. The unique identifier of the package and the creation time; 2. The version number of this encapsulation specification followed; 3. The paths, format descriptions and their roles of the files in the package; 4. The basic metadata of the project (such as project name, filing unit, filing date).

[0063] The archive information package has a clear structure, is self-contained and self-describing. In the future, the structure and content of the entire package can be understood by reading the information package.

[0064] In this embodiment, the archive information package file manifest.json is automatically generated to describe the structure and content of the entire archive package, including: {"@context": "https: / / example.org / bim-archive / v1", "Document ID": "PSP-XC-2024-001", "Creation Time": "2025-09-15T10:00:00Z", "Project Name": "Zhejiang Tiantai Pumped Storage Power Station", "Document": [{"Role": "Geometry", "Path": ". / gltf / transformer_01.gltf", "Format": "gltf+json"}; {"Role": "Attribute", "Path": ". / jsonld / attributes_T-2024-001.jsonld", "Format": "application / ld+json"}; "Role": "Semantics", "Path": ". / rdf / relationships.ttl", "Format": "text / turtle"]}.

[0065] Organize the manifest.json, transformer_01.gltf, attributes_T-2024-001.jsonld, and relationships.ttl files according to the directory structure described in the archive information package.

[0066] All files are packaged into a single file using the standard ZIP compression algorithm and named PSP-XC-2024-001.bima (bima is the extension for BIM Intelligent Model Archive).

[0067] In this embodiment, the self-describing archive package design transforms BIM archives from static storage into understandable, computable, and scalable knowledge assets.

[0068] This embodiment integrates archival value assessment into a lightweight process. By defining LOD archiving levels and corresponding rule bases, it achieves intelligent preservation of core engineering semantic information of electromechanical equipment while reducing data volume, thereby improving information integrity.

[0069] Step S4: Utilization and Verification.

[0070] In the preferred embodiment, the method further includes: querying and reasoning about device attributes and their semantic relationships by parsing the JSON-LD and RDF files in the archive package, supporting intelligent retrieval and knowledge graph construction; all technical parameters of the transformer and its system relationships can be queried.

[0071] Self-describing archives are standard ZIP archives, identified by a specific file extension.

[0072] In this embodiment, users can use any 3D viewer that supports the glTF format (such as Babylon.js, Three.js, Microsoft 3D Viewer) to open the transformer_01.gltf file in the archive package and view the lightweight 3D model of the main transformer.

[0073] Users or applications can parse the manifest.json file, read and load the corresponding JSON-LD and RDF files. Using the SPARQL query language or JSON path, they can precisely retrieve all technical parameters of the main transformer, its system, connection relationships, and associated documentation, thus transforming it from a visual asset into a computable knowledge asset.

[0074] Testing showed that the archive package generated using the method described in this embodiment has a storage size reduced by approximately 75% compared to the original IFC file (this varies depending on the equipment value and LOD level). All data in the archive package is based on W3C open standards and does not rely on any specific BIM software vendor, ensuring long-term readability for over 20 years. Simultaneously, the key geometric features, all technical parameters, and complex system relationships of the main transformer are fully preserved, meeting the archive utilization needs for the entire lifecycle operation and maintenance management of the power plant.

[0075] In summary, the lightweight self-describing encapsulation method for long-term storage of BIM models provided by this invention effectively solves the problems of high storage costs, loss of semantic information, and strong vendor dependence in existing BIM archiving technologies, and has good industrial application prospects and practical value.

[0076] The above embodiments are merely preferred technical solutions of the present invention and should not be considered as limitations on the present invention. The scope of protection of the present invention should be limited to the technical solutions described in the claims, including equivalent substitutions of the technical features described in the claims. That is, equivalent substitutions and improvements within this scope are also within the scope of protection of the present invention.

Claims

1. A lightweight self-describing encapsulation method for long-term storage of BIM models, characterized in that, Includes the following steps: S1: Construct a value assessment model for electromechanical equipment engineering, and quantitatively assess the equipment components in the BIM model based on three dimensions: equipment importance, equipment service life, and semantic demand intensity, and generate a comprehensive equipment value score; S2: Based on the comprehensive value score of the equipment, map the equipment components to the preset LOD archiving level, and perform semantically aware lightweight processing on the BIM model according to the lightweight rule library corresponding to each level to generate a lightweight geometric model and a structured attribute dataset. S3: Organize the lightweight data according to a ternary separation structure, and store geometric information, attribute information and semantic relationship information in open standard formats to form a geometric resource library, an attribute resource library and a semantic relationship library; S4: Construct an archive information package, record the encapsulation specification version, file path, project metadata and file role description, and encapsulate the geometric resource library, attribute resource library, semantic relation library and archive information package together to generate a self-describing archive package.

2. The lightweight self-describing encapsulation method for long-term storage of BIM models according to claim 1, characterized in that, The comprehensive value scoring function of the electromechanical equipment engineering value assessment model is: V = α·I + β·L + γ·S; Where V is the overall value score of the equipment; I is the importance of the equipment; L is the equipment usage cycle; S is the semantic demand intensity; and α, β, and γ are preset weight coefficients.

3. The lightweight self-describing encapsulation method for long-term storage of BIM models according to claim 2, characterized in that, The importance of the equipment, the equipment usage cycle, and the semantic demand intensity are each calculated by weighting multiple sub-indicators, including: The importance of equipment is determined by a weighted average of its impact on system functionality, maintenance difficulty coefficient, and economic value. The equipment service life is obtained by weighting the design life, expected service life, and replacement frequency. The strength of semantic requirements is obtained by weighting the dependence of business systems, the frequency of semantic data usage, and the requirements for information integrity.

4. The lightweight self-describing encapsulation method for long-term storage of BIM models according to claim 1, characterized in that, The LOD archiving levels correspond to different geometric precision and attribute retention granularity, including: LOD Archive 500 preserves the highest precision geometry and all attributes; LOD Archive 400 preserves the main geometric features and key business attributes; LOD Archive 300 retains the basic shape and logo information; LOD Archive 200 preserves system connections and key parameters, and geometrically simplifies them into symbolic representations.

5. The lightweight self-describing encapsulation method for long-term storage of BIM models according to claim 1, characterized in that, The ternary separation structure includes: The geometric resource library uses the glTF 2.0 format to store lightweight 3D geometric information; The attribute resource library uses JSON-LD format to store non-geometric attribute information and supports semantic tags and links to external knowledge bases; A semantic relational database uses RDF format to store the relationships between components and constructs a project knowledge graph.

6. The lightweight self-describing encapsulation method for long-term storage of BIM models according to claim 1, characterized in that, The processing of the geometric resource library includes: adaptive patch simplification based on LOD level, material texture compression, vertex data optimization, similar volume merging, and node hierarchy preservation.

7. The lightweight self-describing encapsulation method for long-term storage of BIM models according to claim 1, characterized in that, The processing of the attribute resource library includes: structured attribute extraction, key attribute filtering, semantic tag annotation, and linking to external knowledge bases.

8. The lightweight self-describing encapsulation method for long-term storage of BIM models according to claim 1, characterized in that, The processing of the semantic relation database includes: component association identification, RDF triple generation, relation ontology definition, and knowledge graph construction.

9. The lightweight self-describing encapsulation method for long-term storage of BIM models according to claim 1, characterized in that, The archive information package includes at least: Unique identifier and creation time of the file package; Packaging specification version number; The path, format, and role description of each resource repository file; Metadata such as project name, archiving unit, and archiving date.

10. The lightweight self-describing encapsulation method for long-term storage of BIM models according to claim 1, characterized in that, The method also includes: parsing JSON-LD and RDF files in the archive package to query and reason about device attributes and their semantic relationships, supporting intelligent retrieval and knowledge graph construction; The self-describing archive is a standard ZIP archive and is identified using a specific file extension.

Images