A graph-based BIM data storage and management method

Abstract

The present invention discloses a graph-based BIM data storage management method, comprising constructing a data structure based on a multi-level graph, wherein the first-level graph is used to represent the overall architecture of different projects in the BIM model, the second-level graph is used to present the structural relationship between different partitions in each project, and the third-level and lower-level graphs are used to carefully depict the entities in the partitions and their mutual relationships; a unique identifier with a timestamp and a version number is assigned to each entity in the BIM data, and the identifier is associated with detailed attribute information including geometric information, material properties, construction technology, etc. and stored; a graph algorithm, such as a graph neural network algorithm based on deep learning, is used to deeply analyze and process the complex relationships in the BIM data to achieve efficient data query, update and prediction operations; the present invention can accurately represent the complex relationships and hierarchical structures in the BIM model by constructing a multi-level graph data structure, facilitates rapid positioning and operation of data, and significantly improves the efficiency of data storage and management.

Description

Technical Field

[0001] The present invention relates to the technical field of data storage management, and more specifically, to a graph-based BIM data storage management method. Background Art

[0002] The application of Building Information Modeling (BIM) technology in the construction industry is becoming more and more widespread. However, it is accompanied by an urgent need for efficient, accurate and secure BIM data storage and management.

[0003] In the existing BIM data processing, traditional data storage methods are often unable to cope with complex and changeable building structures and massive amounts of data. Relational databases show obvious limitations when processing the entity relationships and hierarchical structures of BIM data, and cannot intuitively display the relationship between building elements, resulting in complex and time-consuming data query and update operations.

[0004] As the scale of construction projects continues to expand, the amount of BIM data is growing exponentially. The centralized storage model not only faces storage capacity limitations, but is also prone to single point failures, affecting data availability and reliability.

[0005] For entity identifiers in BIM data, existing methods have problems such as non-unique identification, difficulty in reflecting version changes, and inability to effectively trace data modification history, which poses challenges to data consistency and accuracy.

[0006] In addition, traditional indexing mechanisms are often unable to quickly find the optimal retrieval path when faced with diverse query requirements and complex data features, resulting in low query efficiency and failure to meet the rapid response requirements in practical applications.

[0007] In terms of data security, with the increasing degree of digitization, the sensitive information contained in BIM data faces more and more security threats, and the existing encryption and protection methods are unable to cope with high-level security requirements.

[0008] In summary, traditional BIM data storage and management methods have many shortcomings in data structure representation, storage capacity, identifier management, index efficiency and data security. An innovative graph-based BIM data storage management method is urgently needed to solve these problems. Summary of the invention

[0009] 1. Technical issues to be solved

[0010] The purpose of the present invention is to provide a graph-based BIM data storage and management method to solve the problems raised in the above background technology.

[0011] 2. Technical solution

[0012] A graph-based BIM data storage and management method comprises the following steps:

[0013] S1. Construct a data structure based on a multi-level graph, where the first-level graph is used to represent the overall architecture of different projects in the BIM model, the second-level graph is used to present the structural relationship between different partitions in each project, and the third-level and lower-level graphs are used to carefully depict the entities in the partition and their relationships;

[0014] S2. Assign a unique identifier with a timestamp and version number to each entity in the BIM data, and store it in association with detailed attribute information including geometric information, material properties, and construction process;

[0015] S3. Use graph algorithms, such as graph neural network algorithms based on deep learning, to deeply analyze and process complex relationships in BIM data to achieve efficient data query, update and prediction operations;

[0016] S4. Establish a multi-level indexing mechanism, including primary index based on entity attributes, auxiliary index based on relationship, and dynamic index based on time series, to speed up the retrieval of BIM data;

[0017] S5. Distributed storage technology is used, combined with data sharding strategy and intelligent replica placement algorithm, to distribute and store large-scale BIM data on multiple storage nodes, and a data consistency verification mechanism is used to ensure data consistency and accuracy;

[0018] S6. Introduce a data compression module, use a wavelet transform-based compression algorithm to efficiently compress BIM data, and use an encryption module to encrypt the compressed data using quantum encryption technology to reduce data storage space and ensure data security and privacy;

[0019] S7. Design data migration strategies to automatically migrate data between storage media with different performances based on data access frequency and importance, thus optimizing storage costs and access efficiency.

[0020] S8. Develop data interfaces to support seamless connection with various BIM software and other related systems, realize rapid data import, export and sharing, and equip data format conversion modules to adapt to the data format requirements of different systems;

[0021] S9. Establish a data monitoring and early warning mechanism to monitor data storage and access status in real time, issue timely warnings for abnormal situations, and automatically start data recovery and backup processes.

[0022] Preferably, the data structure of the multi-level graph can accurately and dynamically reflect the changes in the structure and association relationships of different components and elements in the BIM model at different levels and time dimensions.

[0023] Preferably, the generation of the unique identifier with a timestamp and a version number follows strict coding rules, combined with a hash algorithm and digital signature technology to ensure the uniqueness, immutability and traceability of the identifier, wherein the designed coding rule formula is:

[0024] UID=Hash(ProjectID+EntityType+EntityID+VersionNumber+TimeStamp)+DigitalSignature(Hash(ProjectID+EntityType+EntityID+VersionNumber+TimeStamp))

[0025] in:

[0026] ProjectID represents the unique identifier of the project;

[0027] EntityType represents the type identifier of the entity;

[0028] EntityID represents the unique identifier of the entity itself;

[0029] VersionNumber represents the version number of the entity;

[0030] TimeStamp represents the timestamp of generating the identifier;

[0031] The Hash() function uses a secure hash algorithm, such as the SHA-256 algorithm, to hash the input string.

[0032] The DigitalSignature() function signs the hash value using a reliable digital signature algorithm, such as RSA or ECC.

[0033] Preferably, the deep learning-based graph neural network algorithm can automatically learn and extract hidden features and patterns in BIM data, providing more in-depth support for data analysis and decision-making.

[0034] Preferably, the multi-level index mechanism can intelligently select the optimal index path according to different query requirements and data characteristics to improve query efficiency. The designed index path mechanism is:

[0035] Assume that there are three layers of index structure, namely project layer L1, partition layer L2 and entity layer L3;

[0036] Index path design:

[0037] Project layer L1 index: Create a B+ tree index based on the key attributes of the project, such as project name and project number;

[0038] Partition layer L2 index: For each partition in a project, a hash index is created based on the partition characteristics, such as partition type and partition area.

[0039] Entity layer L3 index: For entities in each partition, a composite index is built based on common query attributes of the entity, such as entity name, entity type, and creation time. This is a combination of B+ tree and bitmap index.

[0040] The formula for intelligently selecting the optimal index path is:

[0041] Cost(IndexPath)=C1*QueryComplexity+C2*DataVolume+C3*IndexSize+C4*UpdateFrequency,

[0042] in:

[0043] QueryComplexity: indicates the complexity of the query, such as the number of conditions involved and the complexity of the conditions;

[0044] DataVolume: the corresponding data volume;

[0045] IndexSize: the index size corresponding to the selected index path;

[0046] UpdateFrequency: the frequency of data update;

[0047] C1, C2, C3, and C4 are weight coefficients set according to actual conditions.

[0048] Preferably, the data consistency verification mechanism in the distributed storage technology adopts a distributed hash table and a consensus algorithm to ensure the consistency and accuracy of data on multiple storage nodes.

[0049] Preferably, the compression algorithm based on wavelet transform can adaptively perform compression according to data characteristics, thereby achieving a high compression ratio while ensuring data accuracy.

[0050] Preferably, the quantum encryption technology uses quantum key distribution and quantum encryption algorithm to provide unprecedented security for BIM data.

[0051] Preferably, machine learning algorithms are used to predict data access patterns and trends to achieve intelligent data migration and optimized storage.

[0052] 3. Beneficial effects

[0053] Compared with the prior art, the advantages of the present invention are:

[0054] By constructing a multi-level graph data structure, the present invention can more clearly and accurately represent the complex relationships and hierarchical structures in the BIM model, facilitate rapid positioning and manipulation of data, and significantly improve the efficiency of data storage and management.

[0055] The present invention assigns a unique identifier with a timestamp and version number to each entity, and combines hash algorithm and digital signature technology to ensure the uniqueness, non-tamperability and complete traceability of the data, thereby guaranteeing the accuracy and reliability of the data.

[0056] The innovative multi-level indexing mechanism of the present invention can intelligently select the optimal indexing path according to different query requirements and data characteristics, greatly improving the data retrieval speed and saving query time.

[0057] The distributed storage technology adopted in the present invention combines data sharding strategy, intelligent replica placement algorithm and data consistency verification mechanism to effectively solve the problem of large-scale data storage and ensure the consistency and accuracy of data on multiple storage nodes.

[0058] The present invention adopts advanced wavelet transform-based compression algorithm and quantum encryption technology, which not only reduces data storage space, but also provides high-intensity security protection for BIM data to prevent data leakage and tampering.

[0059] The data migration strategy designed by the present invention can automatically migrate data between storage media with different performances according to the access frequency and importance of the data, thereby optimizing storage costs.

[0060] The data interface developed by the present invention supports seamless connection with a variety of BIM software and related systems, realizes rapid import, export and sharing of data, is conducive to collaborative work between different teams and departments, and improves the overall efficiency of the project.

[0061] The data monitoring and early warning mechanism established by the present invention can monitor the data status in real time, detect and handle abnormal situations in a timely manner, automatically start the data recovery and backup process, and ensure the availability and integrity of the data. At the same time, the machine learning algorithm is used to predict data access patterns and trends, realize intelligent data migration and optimized storage, and further improve the intelligent level of data management. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 It is a schematic block diagram of the graph-based BIM data storage and management method of the present invention. DETAILED DESCRIPTION

[0063] Example: See Figure 1 , a graph-based BIM data storage and management method that covers the following series of carefully designed steps:

[0064] S1. Construct a data structure based on multi-level graphs, where the first-level graph at the top level focuses on depicting the overall architecture of various projects in the BIM model, carefully outlining the macro layout and relationships between projects; the second-level graph focuses on presenting the structural relationship between different partitions within each project, clearly showing the connections and differences between partitions; and the third-level and lower-level graphs are used to meticulously depict the entities within the partitions and their intricate relationships, including the connections, dependencies, and interaction details between entities.

[0065] S2. Carefully assign a unique identifier with a timestamp and version number to each entity in the BIM data, and store it closely associated with detailed and critical attribute information including geometric information, material properties, and construction processes. This design ensures the unique identification of each entity in the data system and a clear and traceable historical change record, providing a solid guarantee for the accuracy and consistency of the data.

[0066] S3. Use advanced graph algorithms, especially graph neural network algorithms based on deep learning, to deeply analyze and process the complex relationships in BIM data. This cutting-edge algorithm can automatically learn and extract hidden features and patterns in BIM data with its powerful learning and mining capabilities, thereby achieving efficient data query, update and accurate prediction operations, providing strong data support and decision-making basis for the planning, execution and management of construction projects.

[0067] S4. Establish a comprehensive and multi-level indexing mechanism, including primary indexes based on entity attributes, auxiliary indexes based on relationships, and dynamic indexes based on time series. Through such a multi-dimensional indexing system, the retrieval speed of BIM data can be greatly accelerated, ensuring that the required information can be quickly located and obtained in massive data, and significantly improving the efficiency and accuracy of data query.

[0068] S5. Use advanced distributed storage technology, cleverly combine data sharding strategy and intelligent copy placement algorithm, and reasonably distribute and store large-scale BIM data on multiple storage nodes. At the same time, use efficient data consistency verification mechanism to continuously monitor and ensure the consistency and accuracy of data between multiple storage nodes, effectively avoid data deviation and error, and ensure data integrity and reliability.

[0069] S6. Introducing a powerful data compression module, using an efficient compression algorithm based on wavelet transform to finely compress BIM data. This advanced compression technology can significantly reduce data storage space while retaining key information and accuracy of the data to the greatest extent. In addition, through the encryption module, the compressed data is strictly encrypted using highly secure quantum encryption technology, thereby effectively ensuring the security and privacy of the data and preventing illegal access and tampering during storage and transmission.

[0070] S7. Carefully design flexible data migration strategies to automatically migrate data between storage media with different performances based on key factors such as data access frequency and importance. Such strategies can optimize storage costs, ensure that high-access frequency and important data are stored on high-performance media to improve access efficiency, and reasonably utilize the characteristics of different storage media to achieve optimal resource allocation.

[0071] S8. Develop highly compatible data interfaces to fully support seamless connection with various BIM software and other related systems, and realize fast data import, export and barrier-free sharing. At the same time, it is equipped with a powerful data format conversion module, which can flexibly adapt to the data format requirements of different systems, eliminate the obstacles caused by data format differences, and promote the smooth flow and efficient use of data between different platforms and systems.

[0072] S9. Establish a rigorous data monitoring and early warning mechanism to monitor the status of data storage and access in real time and continuously. Once an abnormal situation is found, it can immediately issue a timely alarm and automatically start a comprehensive data recovery and backup process. Through this active, real-time monitoring and emergency response mechanism, the availability and integrity of data can be guaranteed to the greatest extent, and the risk of data loss and damage can be reduced.

[0073] In the present invention, the data structure of the multi-level graph has a high degree of accuracy and dynamic adaptability, and can reflect the changes in the structure and relationship of different components and elements in the BIM model at different levels and time dimensions in real time, accurately and dynamically. Whether it is the planning adjustment of the construction project, the changes during the construction process, or the modifications in the later operation and maintenance stage, this data structure can quickly and accurately capture and present these changes, providing a solid data foundation for the full life cycle management of the project.

[0074] In the present invention, the generation of a unique identifier with a timestamp and a version number strictly follows a set of rigorous coding rules, and skillfully combines the hash algorithm and the digital signature technology, thereby effectively ensuring the uniqueness, immutability and complete traceability of the identifier. The carefully designed coding rule formula is: UID = Hash (ProjectID + EntityType + EntityID + VersionNumber + TimeStamp) + DigitalSignature (Hash (ProjectID + EntityType + EntityID + VersionNumber + TimeStamp)) where: ProjectID represents the unique identifier of the project, which is used to clearly distinguish different construction projects; EntityType represents the type identifier of the entity, which clearly defines the category to which the entity belongs, such as beams, columns, and walls; EntityID represents the unique identifier of the entity itself, ensuring the uniqueness of each entity within its type; VersionNumber represents the version number of the entity, recording the previous modifications and updated versions of the entity; TimeStamp represents the precise timestamp for generating the identifier, accurately recording the creation time of the identifier. The Hash() function uses a highly secure hash algorithm, such as the SHA-256 algorithm, which has excellent stability and security, to perform accurate hash calculations on the input string; the DigitalSignature() function uses a reliable digital signature algorithm, such as RSA or ECC, which has extremely high security and credibility, to strictly sign the hash value.

[0075] In this invention, the graph neural network algorithm based on deep learning shows powerful functions, which can automatically learn and deeply extract hidden features and patterns in BIM data, providing more in-depth, comprehensive and accurate support for data analysis and decision-making. By mining the complex relationships and potential laws in the data, it provides valuable reference and guidance for the optimization design, construction progress prediction and cost control of construction projects.

[0076] In the present invention, the multi-level indexing mechanism shows a high degree of intelligence and adaptability, and can flexibly and intelligently select the optimal index path according to different query requirements and unique data characteristics, thereby significantly improving query efficiency. The designed index path mechanism is: assuming that there are three layers of index structure, namely project layer L1, partition layer L2 and entity layer L3; index path design: project layer L1 index: based on the key attributes of the project, such as the core elements of project name and project number, create an efficient B+ tree index to ensure fast retrieval and positioning of project-level data; partition layer L2 index: for each partition in the project, based on the significant features of the partition, such as partition type, partition area key indicators, create an efficient hash index to achieve fast query and screening of partition data; entity layer L3 index: for the entities in each partition, according to the common query attributes of the entity, such as entity name, entity type, creation time important features, build a powerful combined index, cleverly combine B+ tree with bitmap index, and further improve the query speed and accuracy of entity data; intelligently select the optimal index The formula for index path: Cost(IndexPath)=C1*QueryComplexity+C2*DataVolume+C3*IndexSize+C4*UpdateFrequency, where: QueryComplexity indicates the complexity of the query, such as the number of conditions involved and the complexity of the conditions, which directly affects the difficulty and amount of calculation of the query; DataVolume corresponds to the amount of data. The larger the amount of data, the higher the difficulty of query and processing; IndexSize indicates the index size corresponding to the selected index path. An overly large index may affect storage and query efficiency; UpdateFrequency indicates the frequency of data updates. Frequently updated data sets may require different index strategies to ensure performance. C1, C2, C3, and C4 are weight coefficients carefully set according to actual conditions. By reasonably adjusting and optimizing these coefficients, the index selection can be more in line with specific application scenarios and data characteristics, achieving optimal query performance.

[0077] In the present invention, the data consistency verification mechanism in the distributed storage technology adopts advanced distributed hash tables and efficient consensus algorithms to ensure the high consistency and accuracy of data on multiple storage nodes. The distributed hash table is used to quickly locate and distribute data, and the consensus algorithm is combined to reach a consistent state between multiple nodes, thereby effectively avoiding data inconsistency and errors, and ensuring the reliability and stability of the entire distributed storage system.

[0078] In the present invention, the compression algorithm based on wavelet transform has excellent adaptive ability and can flexibly and accurately compress data according to its unique characteristics. Under the strict premise of ensuring that data accuracy is not lost, a remarkably high compression ratio is achieved, which greatly reduces the storage space of data and improves storage efficiency and data transmission speed.

[0079] In this invention, quantum encryption technology uses cutting-edge quantum key distribution and advanced quantum encryption algorithms to provide unprecedented high-level security for BIM data. With the unique properties of quantum physics, the encrypted BIM data is almost impossible to crack, greatly enhancing the confidentiality and integrity of the data and effectively preventing attacks and theft from various potential threats.

[0080] In this invention, advanced machine learning algorithms are used to predict data access patterns and trends, and to achieve intelligent data migration and optimized storage. Through learning and analyzing historical access data, future data access needs can be accurately predicted, and data can be dynamically migrated to the most suitable storage medium, thereby further optimizing storage costs and access efficiency, and improving the performance and flexibility of the entire data storage management system.

[0081] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A graph-based BIM data storage and management method, characterized in that: The following steps are involved: S1. Construct a data structure based on a multi-level graph, where the first-level graph is used to represent the overall architecture of different projects in the BIM model, the second-level graph is used to present the structural relationship between different partitions in each project, and the third-level and lower-level graphs are used to carefully depict the entities in the partition and their relationships; S2. Assign a unique identifier with a timestamp and version number to each entity in the BIM data, and store it in association with information including geometric information, material properties, and construction process properties; S3. Use graph neural network algorithms based on deep learning to deeply analyze and process complex relationships in BIM data to achieve efficient data query, update and prediction operations; S4. Establish a multi-level indexing mechanism, including primary index based on entity attributes, auxiliary index based on relationship, and dynamic index based on time series, to speed up the retrieval of BIM data; S5. Distributed storage technology is used, combined with data sharding strategy and intelligent replica placement algorithm, to distribute and store large-scale BIM data on multiple storage nodes, and a data consistency verification mechanism is used to ensure data consistency and accuracy; S6. Introduce a data compression module, use a wavelet transform-based compression algorithm to efficiently compress BIM data, and use an encryption module to encrypt the compressed data using quantum encryption technology to reduce data storage space and ensure data security and privacy; S7. Design data migration strategies to automatically migrate data between storage media with different performances based on data access frequency and importance, thus optimizing storage costs and access efficiency. S8. Develop data interfaces to support seamless integration with a variety of BIM software and other related systems, realize rapid data import, export and sharing, and equip data format conversion modules to adapt to the data format requirements of different systems; S9. Establish a data monitoring and early warning mechanism to monitor data storage and access status in real time, issue timely warnings for abnormal situations, and automatically start data recovery and backup processes; The generation of the unique identifier with timestamp and version number follows strict coding rules, combined with hash algorithm and digital signature technology, the uniqueness, immutability and traceability of the identifier, wherein the designed coding rule formula is: UID = Hash(ProjectID + EntityType + EntityID + VersionNumber + TimeStamp) + DigitalSignature(Hash(ProjectID + EntityType + EntityID + VersionNumber +TimeStamp)) in: ProjectID represents the unique identifier of the project; EntityType represents the type identifier of the entity; EntityID represents the unique identifier of the entity itself; VersionNumber represents the version number of the entity; TimeStamp represents the timestamp of generating the identifier; The Hash() function uses the SHA-256 algorithm to perform hash calculation on the input string; The DigitalSignature() function uses the digital signature algorithm RSA or ECC to sign the hash value; The deep learning-based graph neural network algorithm can automatically learn and extract hidden features and patterns in BIM data, providing deeper support for data analysis and decision-making; The multi-level index mechanism can intelligently select the optimal index path according to different query requirements and data characteristics to improve query efficiency. The designed index path mechanism is: There are three layers of index structure, namely project layer L1, partition layer L2 and entity layer L3; Index path design: Project layer L1 index: Create a B+ tree index based on the key attributes of the project, including project name and project number; Partition layer L2 index: For each partition in the project, a hash index is created based on the partition characteristics, including partition type and partition area; Entity layer L3 index: For entities in each partition, a composite index is built based on the common query attributes of the entity, including entity name, entity type, and creation time. This is a combination of B+ tree and bitmap index. The formula for intelligently selecting the optimal index path is: Cost(IndexPath) = C1 * QueryComplexity + C2 * DataVolume + C3 * IndexSize + C4 * UpdateFrequency, in: QueryComplexity: indicates the complexity of the query, including the number of conditions involved and the complexity of the conditions; DataVolume: the corresponding data volume; IndexSize: the index size corresponding to the selected index path; UpdateFrequency: the frequency of data update; C1, C2, C3, and C4 are weight coefficients set according to actual conditions; The data consistency verification mechanism in the distributed storage technology adopts a distributed hash table and a consensus algorithm to ensure the consistency and accuracy of data on multiple storage nodes.

2. According to the graph-based BIM data storage and management method of claim 1, it is characterized in that: The data structure of the multi-level diagram can accurately and dynamically reflect the changes in the structure and association relationships of different components and elements in the BIM model at different levels and time dimensions.

3. According to the graph-based BIM data storage and management method of claim 1, it is characterized in that: The compression algorithm based on wavelet transform can adaptively compress data according to data characteristics and achieve a high compression ratio while ensuring data accuracy.

4. According to the graph-based BIM data storage and management method of claim 1, it is characterized in that: The quantum encryption technology uses quantum key distribution and quantum encryption algorithm to provide unprecedented security for BIM data.

5. The method for storing and managing BIM data based on a graph according to claim 1, characterized in that: Use machine learning algorithms to predict data access patterns and trends, and achieve intelligent data migration and optimized storage.

Images