Digital twin application construction method and system supporting collaborative editing and version management

Abstract

This invention belongs to the field of building digital twin applications, and discloses a method and system for building digital twin applications that supports collaborative editing and version management. The method includes: responding to a user's application creation request and generating basic application configuration information; providing a hierarchical and categorized visual component library based on the basic application configuration information, and generating an application page layout by combining user operations and intelligent recommendations; based on the application page layout, performing intelligent collaborative editing based on dynamic granular adjustment and incremental version management based on component dependency graphs to generate application configuration; performing adaptive conversion of multi-source heterogeneous data based on data access requests, and generating a unified format data and application interaction system by combining the application page layout configuration component interaction logic; and performing adaptive switching of rendering modes and GPU resource scheduling based on user hardware configuration, network status, and application running indicators, and generating a digital twin application package by combining the application page layout, application configuration, unified format data, and application interaction system.

Description

Technical Field

[0001] This invention belongs to the technical field of building digital twin applications, and in particular relates to a method and system for building digital twin applications that support collaborative editing and version management. Background Technology

[0002] With the rapid development of the digital economy, digital twin technology has become a core technology connecting the physical and digital worlds, and is widely used in urban governance, industrial manufacturing, transportation, and other fields. However, current digital twin application development still faces many challenges: low efficiency in multi-user collaboration, with existing platforms mostly employing coarse-grained file-level locking, leading to frequent editing conflicts; lack of version management capabilities, failing to automatically save historical versions and support intelligent merging; complex data access and processing, requiring the writing of a large amount of customized code; inefficient rendering resource scheduling, with cloud GPU resource utilization at only around 30%; poor component reusability, and a lack of intelligent recommendation capabilities. These problems result in long development cycles and high costs for digital twin applications, severely restricting the large-scale deployment of the technology.

[0003] Therefore, there is an urgent need for a low-threshold, highly collaborative digital twin application construction method and system with comprehensive version management, intelligent data processing, adaptive rendering, and intelligent component recommendation, in order to break down the application barriers of digital twin technology and promote its large-scale implementation. Summary of the Invention

[0004] This application addresses the problems existing in the prior art by proposing a method and system for building digital twin applications that supports collaborative editing and version management. It solves the problem of multi-user editing conflicts through an intelligent collaborative editing mechanism based on dynamic granularity adjustment, achieves efficient version management through an incremental version storage and intelligent merging mechanism based on component dependency graphs and the longest common subsequence, and optimizes rendering performance through a priority-weighted cloud GPU resource dynamic scheduling algorithm. The application shortens the development cycle of digital twin applications from several months to several days, improves team collaboration efficiency, reduces storage space occupation, and improves the utilization rate of cloud rendering resources.

[0005] To achieve the above objectives, this application provides the following technical solution:

[0006] Firstly, a method for building a digital twin application that supports collaborative editing and version management includes the following steps: S1. Responding to a user's application creation request, initialize the digital twin application project and generate basic application configuration information; S2. Based on the basic application configuration information, provide the user with a hierarchical and categorized visual component library, and generate and integrate the application page layout by combining user operations and intelligent recommendations; S3. Based on the application page layout, perform intelligent collaborative editing based on dynamic granular adjustment and incremental version management based on component dependency graphs to generate the application configuration after multi-person collaboration; S4. Obtain the user's data access request, and perform adaptive conversion of multi-source heterogeneous data based on the data access request, and generate a unified format data and application interaction system by combining the application page layout configuration component interaction logic; S5. Obtain the user's hardware configuration, network status, and application running indicators, perform adaptive switching of rendering mode and GPU resource scheduling based on the user's hardware configuration, network status, and application running indicators, and optimize the application packaging by combining the application page layout, application configuration, unified format data, and application interaction system to generate an independently runnable digital twin application package that supports multiple deployment modes.

[0007] Optionally, in step S2, generating and integrating the application page layout by combining user actions and intelligent recommendations includes:

[0008] S21. Based on the user's component drag-and-drop operation, select components from the component library and add them to the application canvas, adjust the component properties and layout, and generate a preliminary application page layout.

[0009] S22. Based on the user's building behavior and application basic configuration information, perform intelligent business component recommendation and templated automatic assembly based on semantic tags to generate application pages that can be built quickly.

[0010] S23. Integrate the quickly built application pages into the initial application page layout.

[0011] Optionally, in step S3, intelligent collaborative editing based on dynamic granularity adjustment includes:

[0012] S31. Divide application resources into granular editing units with three dimensions: spatial dimension, time dimension, and data dimension. The spatial dimension includes application level, page level, container level, component level, and attribute level.

[0013] S32. Real-time collection of user editing operations, extraction of operation frequency, operation concentration, operation granularity and operation duration features;

[0014] S33. Calculate the current optimal locking granularity level based on the extracted operational features, and dynamically adjust the user's locking range;

[0015] S34. An operation intention prediction model based on long short-term memory network predicts the resource that the user will edit next and pre-locks it;

[0016] S35. An optimistic concurrency control strategy is adopted to detect editing conflicts, and different automatic resolution strategies are used for attribute conflicts, layout conflicts, data conflicts and interaction conflicts.

[0017] Optionally, in step S33, the formula for calculating the optimal locking granularity level is as follows:

[0018]

[0019] Where G is the optimal locking granularity level; f is the operation frequency; c is the operation concentration; g is the operation granularity; and t is the operation duration. , , , These are the weighting coefficients for each feature.

[0020] Optionally, in step S3, incremental version management based on the component dependency graph includes:

[0021] S36. Construct a component dependency graph CDG, wherein the nodes of the component dependency graph include component nodes and data source nodes, and the edges of the component dependency graph include data dependency edges, interaction dependency edges and layout dependency edges.

[0022] S37. When a new version is generated, calculate the difference between the current component dependency graph and the previous version component dependency graph, and only store the difference part;

[0023] S38. Generate a complete snapshot as a checkpoint every N versions;

[0024] S39. During version merging, node conflicts and dependency conflicts are detected based on the component dependency graph. Modifications without logical conflicts are automatically resolved, and modifications with logical conflicts are marked for user confirmation.

[0025] Optionally, in step S4, performing adaptive transformation of multi-source heterogeneous data includes:

[0026] S41 adopts a plug-in architecture to support multiple data source types such as static files, relational databases, NoSQL databases, API interfaces, message queues, industrial protocols, and geospatial data;

[0027] S42. Sample data from the data source and extract data structure, field types, field semantics, and data distribution characteristics;

[0028] S43. Based on the characteristics of the original data and the format requirements of the target components, automatically generate transformation rules including field mapping, type conversion, data cleaning and data aggregation;

[0029] S44. It adopts a real-time data stream processing engine based on Apache Flink, which supports real-time processing of millions of data points per second.

[0030] S45. Assess data quality from four dimensions: completeness, accuracy, consistency, and timeliness. Issue an alarm automatically when the data quality score is below the threshold.

[0031] Optionally, in step S45, the formula for calculating the overall data quality score is as follows:

[0032]

[0033] Where Q is the overall data quality score; C is the completeness score; A is the accuracy score; Co is the consistency score; and T is the timeliness score. , , , These are the weighting coefficients for each dimension.

[0034] Optionally, in step S5, performing adaptive switching of rendering modes and GPU resource scheduling includes:

[0035] S51. Construct a unified abstraction layer for 3D rendering modes, encapsulate local rendering and cloud rendering modes, and provide a unified rendering interface.

[0036] S52. Establish a multi-dimensional user experience quantification model, comprehensively considering frame rate, frame time, loading time, network latency, and image quality score.

[0037] S53. Real-time monitoring of user experience metrics; when the overall user experience score is below the threshold, triggering a rendering mode switching evaluation.

[0038] S54. Employ a multi-attribute decision-making algorithm based on the analytic hierarchy process (AHP) to comprehensively consider user experience, hardware performance, network status, and cost to select the optimal rendering mode.

[0039] S55. Employs a priority-weighted cloud GPU resource dynamic scheduling algorithm to dynamically allocate GPU resources based on the comprehensive priority score of the rendering task;

[0040] S56 employs seamless rendering state transition technology to complete the switching of rendering modes without the user's awareness.

[0041] Optionally, in step S52, the formula for calculating the overall user experience score is as follows:

[0042]

[0043] Among them, QoE is the overall user experience score; FPS is the frame rate score; FT is the frame time score; LT is the loading time score; NL is the network latency score; and QS is the image quality score. , , , , These are the weighting coefficients for each indicator.

[0044] Secondly, the present invention provides a digital twin application building system that supports collaborative editing and version management, for implementing the digital twin application building method that supports collaborative editing and version management as described in the first aspect, comprising:

[0045] The application creation module is used to respond to application creation requests, initialize digital twin application projects, and generate basic application configuration information;

[0046] The page building module provides a hierarchical and categorized visual component library based on the application's basic configuration information, and generates and integrates the application page layout by combining user operations and intelligent recommendations.

[0047] The collaboration and version management module is used to perform intelligent collaborative editing based on dynamic granular adjustment and incremental version management based on component dependency graph, based on the application page layout, and generate the application configuration after multi-person collaboration.

[0048] The data and interaction module is used to respond to data access requests, perform adaptive transformation of multi-source heterogeneous data, combine application page layout configuration component interaction logic, and generate a unified format data and application interaction system.

[0049] The rendering and publishing module is used to adaptively switch rendering modes and schedule GPU resources based on user hardware and network status. It combines application configuration, application page layout, unified format data and application interaction system to package and generate digital twin application packages that support multiple deployment modes.

[0050] The beneficial effects of this application are as follows:

[0051] (1) This application realizes fine-grained resource locking based on the intelligent collaborative editing mechanism of dynamic granularity adjustment, effectively solves the problem of multi-person editing conflict and improves team collaboration efficiency.

[0052] (2) Based on the component dependency graph and the longest common subsequence incremental version storage and intelligent merging mechanism, this application automatically saves all historical versions, improves the success rate of version merging, and reduces the risk of version update.

[0053] (3) The adaptive data conversion engine of this application supports one-click access to multiple data sources and automatically generates conversion rules, reducing the workload of data access.

[0054] (4) This application is based on a priority-weighted cloud GPU resource dynamic scheduling algorithm to improve cloud resource utilization and reduce operating costs. Attached Figure Description

[0055] Figure 1 A schematic diagram illustrating the application scenarios for developing digital twin applications based on existing technologies.

[0056] Figure 2 This is a schematic diagram of a digital twin application construction method that supports collaborative editing and version management, as shown in Embodiment 1 of this application.

[0057] Figure 3 This is a schematic diagram of a digital twin application system that supports collaborative editing and version management, as shown in Embodiment 2 of this application. Detailed Implementation

[0058] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description of this application is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely one preferred embodiment of this application and are only used to explain this application. They do not limit the scope of protection of this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0059] like Figure 1 The diagram illustrates an application scenario for existing digital twin technologies. Existing digital twin applications generally employ a three-tiered linear architecture: a data acquisition terminal, a data processing center, and application terminals. The data acquisition terminal is responsible for collecting raw data from multiple heterogeneous sources such as sensors, business systems, and databases, and transmitting the raw data to the data processing center via the network. The data processing center then cleans, converts, and stores the data before pushing the processing results to various application terminals such as PCs and mobile devices. However, this architecture only achieves basic data flow and fails to address issues such as editing conflicts and chaotic version management in collaborative development. It also lacks intelligent data conversion and rendering resource scheduling capabilities, resulting in long development cycles, high costs, and difficulty in supporting the rapid construction and efficient operation and maintenance of large-scale, complex digital twin applications.

[0060] Example 1:

[0061] like Figure 2 As shown, a method for building a digital twin application that supports collaborative editing and version management includes the following steps:

[0062] S1. In response to the user's application creation request, initialize the digital twin application project and generate basic application configuration information;

[0063] S2. Based on the application's basic configuration information, provide users with a hierarchical and categorized visual component library, and generate and integrate the application page layout by combining user operations and intelligent recommendations;

[0064] S3. Based on the application page layout, perform intelligent collaborative editing based on dynamic granular adjustment and incremental version management based on component dependency graph to generate application configuration after multi-person collaboration;

[0065] S4. Obtain the user's data access request, and perform adaptive transformation of multi-source heterogeneous data based on the data access request. Combine the application page layout configuration component interaction logic to generate a unified format data and application interaction system.

[0066] S5. Obtain user hardware configuration, network status and application running metrics. Based on user hardware configuration, network status and application running metrics, perform adaptive switching of rendering mode and GPU resource scheduling. Combined with application page layout, application configuration, unified format data and application interaction system, optimize application packaging and generate a stand-alone digital twin application package that supports multiple deployment modes.

[0067] In step S1, the system receives application creation requests submitted by users through the web interface. After verifying the user's identity and permissions, a unique application record is created in the database, generating an application ID. A standardized application directory structure is automatically created, including page, component, data, and resource directories, and the application configuration file is initialized, writing parameters such as the user-submitted application name, resolution, adaptation mode, group affiliation, and access permissions into the configuration file. Simultaneously, the system automatically creates the main branch and development branch, generating an initial version. This standardized application initialization process ensures that all applications have a unified structure and configuration, laying the foundation for subsequent collaborative development and version management, and shortening project initialization time.

[0068] In step S2, suitable components are selected from the component library based on the industry and scenario tags in the application's basic configuration information. The component library adopts a hierarchical classification design, with at least six categories: basic UI components, 2D geographic information components, 3D twin scene components, layout container components, data visualization components, and reusable business components. Each component has predefined standardized interfaces, including attribute interfaces for configuring appearance and behavior, data interfaces for receiving and processing data, and interaction event interfaces for triggering and responding to interactions. The selected components are displayed visually in the component library panel for users to drag and drop. This hierarchical component library design allows users to quickly locate the required components, and the standardized interface design ensures compatibility and interoperability between components, improving component reusability.

[0069] Specifically, in step S2, generating and integrating the application page layout by combining user actions and intelligent recommendations includes:

[0070] S21. Based on the user's component drag-and-drop operation, select components from the component library and add them to the application canvas, adjust the component properties and layout, and generate a preliminary application page layout.

[0071] S22. Based on the user's building behavior and application basic configuration information, perform intelligent business component recommendation and templated automatic assembly based on semantic tags to generate application pages that can be built quickly.

[0072] S23. Integrate the quickly built application pages into the initial application page layout.

[0073] Users select components from the component library and add them to the application canvas via drag-and-drop. Component instances are created on the canvas, generating unique component IDs and adding them to the page's component tree. Users can adjust component attributes such as size, position, and color through the property configuration panel, and adjust styles such as borders, shadows, and backgrounds through the style configuration panel. Drag-and-drop operations also allow for adjustments to component layout and hierarchy. User modifications are rendered in real-time, providing a WYSIWYG development experience. After completing the initial page design, users can save the page configuration information and generate a preliminary application page layout. This visual drag-and-drop operation eliminates the need for coding, significantly lowering the development threshold and enabling non-professional developers to quickly build application pages, improving page creation efficiency.

[0074] In step S22, the intelligent business component recommendation and templated automatic assembly based on semantic tags includes:

[0075] S221. Construct a multi-dimensional semantic tagging system for business components, including functional tags, industry tags, scenario tags, visual tags, and data tags;

[0076] S222. Collect user building behavior sequences and use an LSTM-based behavior sequence analysis model to identify user building intentions;

[0077] S223. A hybrid recommendation algorithm based on collaborative filtering and content matching is adopted to recommend suitable business components to users; the calculation formula of the hybrid recommendation algorithm is shown below:

[0078]

[0079] in, The recommended score is a composite score. The score is based on content matching and reflects the similarity between the component and the current application context. The score is based on collaborative filtering and reflects the frequency of use of this component by similar users. represents the weighting coefficient. The system sorts the components based on their final scores and recommends the top-N components with the highest scores to the user.

[0080] S224 provides a large number of industry-standard page templates and supports automatic page assembly based on templates. Users only need to make minor adjustments to complete page construction.

[0081] By using a hybrid recommendation algorithm to improve recommendation accuracy, the time users spend searching for components is significantly reduced; and by using automatic template assembly technology, page build time is shortened and component reuse rate is improved.

[0082] In step S3, intelligent collaborative editing based on dynamic granularity adjustment includes:

[0083] S31. Divide application resources into granular editing units with three dimensions: spatial dimension, time dimension, and data dimension. The spatial dimension includes application level, page level, container level, component level, and attribute level; the time dimension includes long-term locking, short-term locking, and instantaneous locking; and the data dimension includes data source level, interface level, and field level.

[0084] S32. Real-time collection of user editing operations, extraction of operation frequency, operation concentration, operation granularity and operation duration features.

[0085] S33. Calculate the current optimal locking granularity level based on the extracted operational features, and dynamically adjust the user's locking range; the calculation formula for the optimal locking granularity level is as follows:

[0086]

[0087] Where G is the current optimal locking granularity level; f is the operation frequency, i.e. the number of operations per unit time; c is the operation concentration, i.e. the proportion of consecutive operations on the same editing unit to the total number of operations; g is the operation granularity, i.e. the average granularity level of the operation object; and t is the operation duration, i.e. the average duration of a single operation. , , , These are the weighting coefficients for each feature.

[0088] S34. An operation intention prediction model based on a long short-term memory network predicts the resource that the user will edit next and pre-locks it. The pre-lock validity period is 10 seconds.

[0089] S35. An optimistic concurrency control strategy is adopted to detect editing conflicts, and different automatic resolution strategies are used for attribute conflicts, layout conflicts, data conflicts and interaction conflicts.

[0090] Through the above steps, the dynamic granularity adjustment mechanism enables fine-grained resource locking and improves line editing capabilities; the pre-locking mechanism reduces the average locking waiting time; and the automatic resolution algorithm can handle more than 95% of editing conflicts, reducing manual intervention and improving team collaboration efficiency.

[0091] In step S3, incremental version management based on the component dependency graph includes:

[0092] S36. Construct a component dependency graph (CDG). The nodes of the component dependency graph include component nodes and data source nodes. The edges of the component dependency graph include data dependency edges, interaction dependency edges, and layout dependency edges. The system automatically maintains the CDG during the application construction process.

[0093] S37. When a new version is generated, calculate the difference between the current component dependency graph CDG and the previous version component dependency graph CDG, and store only the difference part instead of the complete component dependency graph CDG.

[0094] S38. Generate a complete snapshot as a checkpoint every N versions to improve version rebuild efficiency;

[0095] S39. During version merging, node conflicts and dependency conflicts are detected based on the component dependency graph CDG. Modifications without logical conflicts are automatically resolved, and modifications with logical conflicts are marked for user confirmation. Version difference calculation uses the longest common subsequence (LCS) basis, and the specific formula is as follows:

[0096] LCS(i,j)=0, when i=0 or j=0

[0097] LCS(i,j)=LCS(i-1,j-1)+1, when a i =b j hour

[0098] LCS(i,j)=max(LCS(i-1,j),LCS(i,j-1)), when a i ≠b j hour

[0099] Where A=[a1,a2,...,a] m ] and B=[b1,b2,...,b n [m,n] represents the component sequences of two versions, and LCS(m,n) is the length of their longest common subsequence. The system automatically resolves modifications without logical conflicts and marks modifications with logical conflicts for user confirmation. Incremental storage algorithms reduce version storage space usage; LCS-based difference calculation improves accuracy; and a dependency-aware intelligent merging mechanism increases the success rate of version merging and shortens merging time.

[0100] In step S4, performing adaptive transformation of multi-source heterogeneous data includes:

[0101] S41 adopts a plug-in architecture to support multiple data source types such as static files, relational databases, NoSQL databases, API interfaces, message queues, industrial protocols, and geospatial data;

[0102] S42. Sample data from the data source and extract data structure, field types, field semantics, and data distribution characteristics;

[0103] S43. Based on the characteristics of the original data and the format requirements of the target components, automatically generate transformation rules including field mapping, type conversion, data cleaning and data aggregation;

[0104] S44. It adopts a real-time data stream processing engine based on Apache Flink, which supports real-time processing of millions of data points per second with latency controlled within 100 milliseconds.

[0105] S45. Data quality is assessed from four dimensions: completeness, accuracy, consistency, and timeliness. An alert is automatically issued when the data quality score falls below a threshold. The formula for calculating the overall data quality score is as follows:

[0106]

[0107] Where Q is the overall data quality score; C is the completeness score, reflecting whether there are missing values ​​in the data; A is the accuracy score, reflecting whether the data is accurate and error-free; Co is the consistency score, reflecting whether the data is consistent across different data sources; and T is the timeliness score, reflecting whether the data is updated in a timely manner. , , , These are the weighting coefficients for each dimension. The system automatically issues an alarm when q < 60. The plug-in access framework supports almost all common data source types, shortening data access time; the adaptive transformation rule generation algorithm reduces data processing workload and lowers the data processing error rate.

[0108] In step S5, the user selects the triggering component and triggering event (such as button click or dropdown selection) through a visual interactive configuration interface, then selects the response component and response action, and finally configures the action parameters. The system automatically generates the interaction logic code and saves it to the application configuration. Supported interaction logic includes at least data linkage (a change in the data of one component causes other components to update automatically), component showing / hiding (showing or hiding components based on conditions), page navigation (jumping from the current page to another page), pop-up display (displaying detailed information in a modal window), and calling third-party interfaces (calling external APIs to obtain data or perform operations).

[0109] Specifically, in step S5, the adaptive switching of rendering modes and GPU resource scheduling include:

[0110] S51. Construct a unified abstraction layer for 3D rendering modes, encapsulate local rendering and cloud rendering modes, and provide a unified rendering interface; through adaptive rendering mode switching, it ensures a smooth experience under different hardware configurations. Local rendering can give full play to the computing performance of user devices, while cloud rendering can ensure the running effect of low-configuration terminals, thereby improving the overall user experience score.

[0111] S52. Establish a multi-dimensional user experience quantification model, comprehensively considering frame rate, frame time, loading time, network latency, and image quality score; the formula for calculating the comprehensive user experience score is as follows:

[0112]

[0113] Among them, QoE is the overall user experience score; FPS is the frame rate score; FT is the frame time score; LT is the loading time score; NL is the network latency score; and QS is the image quality score. , , , , These are the weighting coefficients for each indicator.

[0114] S53. Real-time monitoring of user experience metrics; when the overall user experience score is below the threshold, triggering a rendering mode switching evaluation.

[0115] S54. Employ a multi-attribute decision algorithm based on the Analytic Hierarchy Process (AHP) to comprehensively consider user experience, hardware performance, network status, and cost to select the optimal rendering mode.

[0116] S55. A priority-weighted cloud GPU resource dynamic scheduling algorithm is adopted to dynamically allocate GPU resources according to the comprehensive priority score of the rendering task; the calculation formula for the GPU resource allocation ratio is as follows:

[0117]

[0118] in, The proportion of GPU resources allocated to the i-th view rendering task; The overall priority score is the score for the j-th parallel view rendering task; j is the summation variable; n is the total number of current parallel view rendering tasks with the overall priority score.

[0119] S56 employs seamless rendering state transition technology, completing the switching of rendering modes within 1 second without the user noticing.

[0120] Local rendering refers to rendering computations being performed entirely on the user's terminal device. It can be implemented using various technologies such as WebGL, WebGPU, and native client graphics interfaces, without relying on cloud computing resources.

[0121] In step S5, during packaging optimization, after receiving the user's release command, the following packaging optimization operations are performed: code compression (compressing and obfuscating JavaScript, CSS, and other code to reduce file size), resource optimization (compressing and optimizing resources such as images and models to improve loading speed), dependency packaging (packaging all dependency libraries together to avoid missing dependencies), and environment configuration (configuring application parameters according to the deployment environment). The system generates a standalone application package, supporting three modes: private deployment (deployed to the user's local server), cloud deployment (deployed to a cloud platform), and hybrid deployment (partial local deployment and partial cloud deployment). A deployment wizard is provided to guide users through the deployment process. The keyless packaging and deployment function significantly simplifies the application release process, reducing application launch time from an average of one week to one hour.

[0122] Example 2:

[0123] like Figure 3 As shown, this embodiment provides a digital twin application building system that supports collaborative editing and version management, used to implement the digital twin application building method supporting collaborative editing and version management as described in Embodiment 1, including:

[0124] The application creation module is used to respond to application creation requests, initialize digital twin application projects, and generate basic application configuration information;

[0125] The page building module provides a hierarchical and categorized visual component library based on the application's basic configuration information, and generates and integrates the application page layout by combining user operations and intelligent recommendations.

[0126] The collaboration and version management module is used to perform intelligent collaborative editing based on dynamic granular adjustment and incremental version management based on component dependency graph, based on the application page layout, and generate the application configuration after multi-person collaboration.

[0127] The data and interaction module is used to respond to data access requests, perform adaptive transformation of multi-source heterogeneous data, combine application page layout configuration component interaction logic, and generate a unified format data and application interaction system.

[0128] The rendering and publishing module is used to adaptively switch rendering modes and schedule GPU resources based on user hardware and network status. It combines application configuration, application page layout, unified format data and application interaction system to package and generate digital twin application packages that support multiple deployment modes.

[0129] The above-described specific embodiments are preferred embodiments of a digital twin application construction method and system that supports collaborative editing and version management according to this application. They are not intended to limit the specific implementation scope of this application. The scope of this application includes but is not limited to the specific embodiments described above. All equivalent changes made in accordance with the shape and structure of this application are within the protection scope of this application.

Claims

1. A method for building a digital twin application that supports collaborative editing and version management, characterized in that, Includes the following steps: S1. In response to the user's application creation request, initialize the digital twin application project and generate basic application configuration information; S2. Based on the application's basic configuration information, provide users with a hierarchical and categorized visual component library, and generate and integrate the application page layout by combining user operations and intelligent recommendations; S3. Based on the application page layout, perform intelligent collaborative editing based on dynamic granularity adjustment and incremental version management based on component dependency graph to generate the application configuration after multi-person collaboration; S4. Obtain the user's data access request, and perform multi-source heterogeneous data adaptive conversion based on the data access request. Combine the application page layout configuration component interaction logic to generate a unified format data and application interaction system. S5. Obtain user hardware configuration, network status and application running indicators. Based on user hardware configuration, network status and application running indicators, perform adaptive switching of rendering mode and GPU resource scheduling. Combine the application page layout, application configuration, unified format data and application interaction system to optimize application packaging and generate a stand-alone digital twin application package that supports multiple deployment modes.

2. The method for building a digital twin application supporting collaborative editing and version management according to claim 1, characterized in that, In step S2, the process of generating and integrating the application page layout by combining user operations and intelligent recommendations includes: S21. Based on the user's component drag-and-drop operation, select components from the component library and add them to the application canvas, adjust the component properties and layout, and generate a preliminary application page layout. S22. Based on the user's building behavior and application basic configuration information, perform intelligent business component recommendation and templated automatic assembly based on semantic tags to generate application pages that can be built quickly. S23. Integrate the quickly built application pages into the initial application page layout.

3. The method for building a digital twin application supporting collaborative editing and version management according to claim 1, characterized in that, In step S3, the intelligent collaborative editing based on dynamic granularity adjustment includes: S31. Divide application resources into granular editing units with three dimensions: spatial dimension, time dimension, and data dimension. The spatial dimension includes application level, page level, container level, component level, and attribute level. S32. Real-time collection of user editing operations, extraction of operation frequency, operation concentration, operation granularity and operation duration features; S33. Calculate the current optimal locking granularity level based on the extracted operational features, and dynamically adjust the user's locking range; S34. An operation intention prediction model based on long short-term memory network predicts the resource that the user will edit next and pre-locks it; S35. An optimistic concurrency control strategy is adopted to detect editing conflicts, and different automatic resolution strategies are used for attribute conflicts, layout conflicts, data conflicts and interaction conflicts.

4. The method for building a digital twin application supporting collaborative editing and version management according to claim 3, characterized in that, In step S33, the calculation formula for the optimal locking granularity level is as follows: Where G is the optimal locking granularity level; f is the operation frequency; c is the operation concentration; g is the operation granularity; and t is the operation duration. , , , These are the weighting coefficients for each feature.

5. The method for building a digital twin application supporting collaborative editing and version management according to claim 1, characterized in that, In step S3, the incremental version management based on the component dependency graph includes: S36. Construct a component dependency graph CDG, wherein the nodes of the component dependency graph include component nodes and data source nodes, and the edges of the component dependency graph include data dependency edges, interaction dependency edges and layout dependency edges. S37. When a new version is generated, calculate the difference between the current component dependency graph and the previous version component dependency graph, and only store the difference part; S38. Generate a complete snapshot as a checkpoint every N versions; S39. During version merging, node conflicts and dependency conflicts are detected based on the component dependency graph. Modifications without logical conflicts are automatically resolved, and modifications with logical conflicts are marked for user confirmation.

6. The method for building a digital twin application supporting collaborative editing and version management according to claim 1, characterized in that, In step S4, the adaptive transformation of multi-source heterogeneous data includes: S41 adopts a plug-in architecture to support multiple data source types such as static files, relational databases, NoSQL databases, API interfaces, message queues, industrial protocols, and geospatial data; S42. Sample data from the data source and extract data structure, field types, field semantics, and data distribution characteristics; S43. Based on the characteristics of the original data and the format requirements of the target components, automatically generate transformation rules including field mapping, type conversion, data cleaning and data aggregation; S44. It adopts a real-time data stream processing engine based on Apache Flink, which supports real-time processing of millions of data points per second. S45. Assess data quality from four dimensions: completeness, accuracy, consistency, and timeliness. Issue an alarm automatically when the data quality score is below the threshold.

7. The method for building a digital twin application supporting collaborative editing and version management according to claim 6, characterized in that, In step S45, the formula for calculating the comprehensive data quality score is as follows: Where Q is the overall data quality score; C is the completeness score; A is the accuracy score; Co is the consistency score; and T is the timeliness score. , , , These are the weighting coefficients for each dimension.

8. The method for building a digital twin application supporting collaborative editing and version management according to claim 1, characterized in that, In step S5, the adaptive switching of rendering modes and GPU resource scheduling includes: S51. Construct a unified abstraction layer for 3D rendering modes, encapsulate local rendering and cloud rendering modes, and provide a unified rendering interface. S52. Establish a multi-dimensional user experience quantification model, comprehensively considering frame rate, frame time, loading time, network latency, and image quality score. S53. Real-time monitoring of user experience metrics; when the overall user experience score is below the threshold, triggering a rendering mode switching evaluation. S54. Employ a multi-attribute decision-making algorithm based on the analytic hierarchy process (AHP) to comprehensively consider user experience, hardware performance, network status, and cost to select the optimal rendering mode. S55. Employs a priority-weighted cloud GPU resource dynamic scheduling algorithm to dynamically allocate GPU resources based on the comprehensive priority score of the rendering task; S56 employs seamless rendering state transition technology to complete the switching of rendering modes without the user's awareness.

9. The method for building a digital twin application supporting collaborative editing and version management according to claim 8, characterized in that, In step S52, the formula for calculating the comprehensive user experience score is as follows: Among them, QoE is the overall user experience score; FPS is the frame rate score; FT is the frame time score; LT is the loading time score; NL is the network latency score; and QS is the image quality score. , , , , These are the weighting coefficients for each indicator.

10. A digital twin application building system that supports collaborative editing and version management, characterized in that, A method for building a digital twin application supporting collaborative editing and version management as described in any one of claims 1-9, comprising: The application creation module is used to respond to application creation requests, initialize digital twin application projects, and generate basic application configuration information; The page building module is used to provide a hierarchical and categorized visual component library based on the application's basic configuration information, and to generate and integrate the application page layout by combining user operations and intelligent recommendations. The collaboration and version management module is used to perform intelligent collaborative editing based on dynamic granular adjustment and incremental version management based on component dependency graph, based on the application page layout, and generate the application configuration after multi-person collaboration. The data and interaction module is used to respond to data access requests, perform adaptive conversion of multi-source heterogeneous data, and generate a unified format data and application interaction system by combining the interaction logic of the application page layout configuration components. The rendering and publishing module is used to perform adaptive switching of rendering modes and GPU resource scheduling based on user hardware and network status. Combined with the application configuration, application page layout, unified format data and application interaction system, it packages and generates a digital twin application package that supports multiple deployment modes.

Images