CN-121543182-B - Model quality evaluation method, device, electronic equipment, product and storage medium

CN121543182BCN 121543182 BCN121543182 BCN 121543182BCN-121543182-B

Abstract

The embodiment of the application discloses a model quality evaluation method, a device, electronic equipment, a product and a storage medium, wherein the method comprises the steps of obtaining a source file of a building information model and engineering construction specification information corresponding to the building information model; extracting association relation between at least one canonical entity and canonical entity from engineering construction canonical information to construct a target knowledge graph, extracting information associated with at least one building element in a building information model from a source file to obtain building element association information, converting the building element association information into knowledge structured data matched with a data format of the target knowledge graph, and evaluating the quality of the building information model according to the knowledge structured data and the target knowledge graph to obtain an evaluation result. According to the scheme provided by the application, the efficiency and the accuracy of the quality evaluation result of the building information model can be improved.

Inventors

ZHANG XIAOCHUN
TENG YIFAN
CHEN WEIDI
LI ZIXUAN
LIU JIE

Assignees

深圳市智慧城市科技发展集团有限公司
深圳市智城软件技术服务有限公司

Dates

Publication Date: 20260508
Application Date: 20260120

Claims (12)

1. A method for evaluating model quality, the method comprising: acquiring a source file of a building information model and engineering construction specification information corresponding to the building information model; extracting at least one canonical entity from the engineering construction canonical information and an association relation between the canonical entities to construct a target knowledge graph; extracting information related to at least one building component in the building information model from the source file to obtain building component related information; Converting the building element association information into knowledge structured data matched with the data format of the target knowledge graph; According to the knowledge structured data and the target knowledge graph, evaluating the quality of the building information model to obtain an evaluation result; before extracting the information associated with at least one building element in the building information model in the source file, the method further comprises: Different data cleaning flows are executed aiming at files with different source formats in the source file so as to remove invalid redundant data and repair broken data chains, and first preprocessed data are obtained, wherein the first preprocessed data comprise RVT data, DWG data and 3DM data; preserving parameterized attributes of the RVT data, strengthening mapping association of two-dimensional labels and a three-dimensional model in the DWG data, and optimizing a conversion algorithm of curved surface data in the 3DM data to obtain second preprocessed data; converting the second preprocessed data into a preset data format through a format conversion tool to obtain third preprocessed data; and comparing the source file with the third preprocessed data, and taking the third preprocessed data as converted data after confirming that the number of components, key attributes and spatial topological relation in the third preprocessed data are not lost or tampered, wherein the converted data are used for acquiring the building component association information.
2. The model quality evaluation method according to claim 1, wherein the extracting information associated with at least one building element in the building information model in the source file to obtain building element associated information includes: Extracting component information of a building information model from the source file, wherein the component information comprises component attributes, component space relations, component material information and component connection relations; extracting multidimensional associated information corresponding to the component information from the source file, wherein the multidimensional associated information comprises an industry specification clause serial number, a design stage version number, a construction process requirement, material provider qualification information and quality detection records of similar components of the past engineering corresponding to the component; And integrating the component information with the multidimensional association information to obtain the building component association information.
3. The model quality evaluation method according to claim 2, wherein the converting the building element association information into knowledge-structured data matching a data format of the target knowledge-graph includes: adding a unique identification code for each data node in the building component association information, and establishing a mapping relation among the component information, the multidimensional association information and the unique identification code to obtain initial knowledge structured data; and converting the initial knowledge structured data into a resource description framework or a network ontology language format to obtain knowledge structured data matched with the data format of the target knowledge graph.
4. The model quality evaluation method of claim 1, wherein extracting an association relationship between at least one canonical entity and the canonical entity in the engineering construction canonical information to construct a target knowledge graph comprises: Acquiring a specification document from the engineering construction specification information, wherein the specification document comprises structured data, semi-structured data and unstructured data; Knowledge extraction and labeling are carried out on the canonical document so as to identify canonical terms, canonical definitions and canonical relationships, and the canonical terms, the canonical definitions and the canonical relationships are converted into structured canonical texts; extracting component types, component attributes and relationships among components from the specification text, and associating the component types, component attributes and relationships among components with corresponding definitions and attributes to obtain specification entities and association relationships among the specification entities; And taking the standard entities as nodes of a target knowledge graph, and taking the association relation between the standard entities as edges of the target knowledge graph to obtain the target knowledge graph.
5. The model quality evaluation method according to claim 1, further comprising, after the constructing the target knowledge graph: determining mandatory levels corresponding to canonical entities in the target knowledge graph according to the canonical attributes; determining engineering key degrees corresponding to the canonical entities in the target knowledge graph according to the engineering stage; determining the execution priority of each canonical entity based on the mandatory level and the engineering criticality; And establishing an index table corresponding to the target knowledge graph according to the execution priority, so that the canonical entity is positioned and called through the index table when the quality of the building information model is evaluated.
6. The model quality evaluation method according to claim 1, wherein the evaluating the quality of the building information model based on the knowledge-structured data and the target knowledge-graph comprises: Checking the missing condition of components and the missing condition of associated data in the building information model according to the knowledge structured data and the target knowledge graph; Checking the specification compliance of the building information model according to the knowledge structured data and the target knowledge graph; Checking the missing condition of component attributes and the accuracy of parameters in the building information model according to the knowledge structured data and the target knowledge graph; And obtaining a quality evaluation result of the building information model based on the missing condition of the component, the missing condition of the associated data, the standard conforming condition, the missing condition of the component attribute and the accuracy of the parameter.
7. The model quality evaluation method according to claim 1, wherein after the quality of the building information model is evaluated, further comprising: Determining illegal component information according to the spatial topological relation of the target knowledge graph and a data association network, wherein the illegal component information is a three-dimensional coordinate, a professional partition and an association component group of a component which does not accord with specifications in the building information model; positioning a data error source by tracing the association relation of the data nodes in the knowledge structured data; And generating a model detection report according to the violation component information and the data error source.
8. The model quality assessment method according to claim 7, further comprising, after said generating a model detection report: Determining a modification scheme corresponding to each non-normative component in the building information model based on the correct case in the target knowledge graph, industry best practice data and associated construction process requirements; and sequencing the modification schemes by combining the priority of the specifications in the target knowledge graph and the influence range of the components which do not meet the specifications to obtain modification priorities, so that the modification schemes are executed based on the modification priorities.
9. A model quality evaluation device, characterized in that the device comprises: the information acquisition unit is used for acquiring a source file of the building information model and engineering construction specification information corresponding to the building information model; The knowledge graph construction unit is used for extracting at least one canonical entity from the engineering construction canonical information and the association relation between the canonical entities so as to construct a target knowledge graph; a building element association information obtaining unit, configured to extract information associated with at least one building element in the building information model from the source file, to obtain building element association information; the information conversion unit is used for converting the building element association information into knowledge structured data matched with the data format of the target knowledge graph; The quality evaluation unit is used for evaluating the quality of the building information model according to the knowledge structural data and the target knowledge graph to obtain an evaluation result; before extracting the information associated with at least one building element in the building information model in the source file, the method further comprises: Different data cleaning flows are executed aiming at files with different source formats in the source file so as to remove invalid redundant data and repair broken data chains, and first preprocessed data are obtained, wherein the first preprocessed data comprise RVT data, DWG data and 3DM data; preserving parameterized attributes of the RVT data, strengthening mapping association of two-dimensional labels and a three-dimensional model in the DWG data, and optimizing a conversion algorithm of curved surface data in the 3DM data to obtain second preprocessed data; converting the second preprocessed data into a preset data format through a format conversion tool to obtain third preprocessed data; and comparing the source file with the third preprocessed data, and taking the third preprocessed data as converted data after confirming that the number of components, key attributes and spatial topological relation in the third preprocessed data are not lost or tampered, wherein the converted data are used for acquiring the building component association information.
10. An electronic device, comprising a processor and a memory, wherein the memory stores a plurality of instructions, and the processor loads instructions from the memory to perform the steps in the model quality evaluation method according to any one of claims 1 to 8.
11. A computer program product comprising computer programs/instructions which, when executed by a processor, implement the steps of the model quality assessment method according to any one of claims 1 to 8.
12. A computer readable storage medium, characterized in that it stores a plurality of instructions adapted to be loaded by a processor for executing the steps of the model quality assessment method according to any one of claims 1-8.

Description

Model quality evaluation method, device, electronic equipment, product and storage medium Technical Field The application relates to the technical field of data processing, in particular to a model quality evaluation method, a device, electronic equipment, a product and a storage medium. Background BIM (Building Information Model ) is a digital carrier integrating multidimensional information such as geometric shape, material performance, construction process, cost progress and the like in the whole life cycle of building projects, and can realize cooperation, visual simulation and data sharing of each specialty. The BIM model quality evaluation is used for guaranteeing the application value, solving the problems through professional software verification, manual rechecking, collaborative examination and other modes around the development of the accuracy, consistency, integrity and normalization of the model, ensuring that the model can reliably support the follow-up construction guidance, cost accounting, operation and maintenance management and other works, and avoiding the actual engineering risks caused by model defects. However, in existing BIM model quality evaluation, a prestored knowledge base is mostly utilized to search for a model collision quality problem, and rule base covered by a BIM model quality automatic evaluation rule knowledge base has limited content, so that requirement information of an evaluation problem needs to be acquired first, and then the requirement information of the evaluation problem is translated into a language recognizable by a microcomputer, so that efficiency and accuracy of a model quality evaluation result are reduced. Therefore, how to improve the efficiency and accuracy of the BIM quality evaluation result is a problem that needs to be solved at present. Disclosure of Invention The embodiment of the application provides a model quality evaluation method, a device, electronic equipment, a product and a storage medium, which can improve the efficiency and accuracy of BIM quality evaluation results. The embodiment of the application provides a model quality evaluation method, which comprises the following steps: acquiring a source file of a building information model and engineering construction specification information corresponding to the building information model; extracting at least one canonical entity from the engineering construction canonical information and an association relation between the canonical entities to construct a target knowledge graph; extracting information related to at least one building component in the building information model from the source file to obtain building component related information; Converting the building element association information into knowledge structured data matched with the data format of the target knowledge graph; and evaluating the quality of the building information model according to the knowledge structured data and the target knowledge graph to obtain an evaluation result. In some embodiments, before extracting information associated with at least one building element in the building information model in the source file, the method further comprises: Different data cleaning flows are executed aiming at files with different source formats in the source file so as to remove invalid redundant data and repair broken data chains, and first preprocessed data are obtained, wherein the first preprocessed data comprise RVT data, DWG data and 3DM data; preserving parameterized attributes of the RVT data, strengthening mapping association of two-dimensional labels and a three-dimensional model in the DWG data, and optimizing a conversion algorithm of curved surface data in the 3DM data to obtain second preprocessed data; converting the second preprocessed data into a preset data format through a format conversion tool to obtain third preprocessed data; and comparing the source file with the third preprocessed data, and taking the third preprocessed data as converted data after confirming that the number of components, key attributes and spatial topological relation in the third preprocessed data are not lost or tampered, wherein the converted data are used for acquiring the building component association information. In some embodiments, the extracting information associated with at least one building element in the building information model in the source file to obtain building element associated information includes: Extracting component information of a building information model from the source file, wherein the component information comprises component attributes, component space relations, component material information and component connection relations; extracting multidimensional associated information corresponding to the component information from the source file, wherein the multidimensional associated information comprises an industry specification clause serial number, a design stage version number,