Electronic sand table personalized recommendation system in a virtual construction system based on graph rag

Abstract

The application discloses a kind of electronic sand table personalized recommendation system in virtual construction system based on GraphRAG, it includes: data acquisition module: for collecting user behavior data, scene perception data and industry knowledge data;Data processing and analysis module: for analyzing user behavior data and scene identification and comparison;GraphRAG retrieval enhancement module: for knowledge graph construction, template and component library retrieval and enhance retrieval capability;Personalized recommendation module: for sand table template and sand table component recommendation;Personalized explanation generation module: using GraphRAG technology, for the template and component of recommended generation personalized explanation;Visual interactive module: including template, component interaction and dialogue interaction;Dynamic adjustment update module: for scene change identification and user feedback analysis.The application is realized scene perception and personalized recommendation by combining the behavior data of user and three-dimensional scene information, using industry knowledge graph, effectively improve the intelligent level of virtual construction.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of virtual construction and intelligent recommendation, and particularly relates to an electronic sand table personalized recommendation system in a virtual construction system based on GraphRAG. BACKGROUND

[0002] In a virtual construction system, users often need to create various three-dimensional scenes according to different design requirements, and select suitable sand table templates and components for design and display. However, existing virtual construction systems usually rely on static template libraries and component libraries, lacking intelligent perception and dynamic response capabilities for user-specific needs and behaviors. Users often feel confused when faced with a large number of template and component choices, making it difficult to quickly find the most suitable design solution, thereby affecting design efficiency and quality.

[0003] In addition, with the complication of virtual construction application scenarios, such as transportation infrastructure, urban planning, environmental protection, etc., user design requirements have become more diverse. Existing technologies are difficult to fully utilize user behavior data and scene information for effective scene recognition and personalized recommendation. This results in virtual construction systems lacking in pertinence and intelligence when applied to complex scenarios, failing to meet users' efficient design needs.

[0004] The existing technology in the virtual construction system faces the following main problems, which limit the intelligence and efficiency of the system in complex design scenarios:

[0005] 1. Insufficient scene recognition and perception capabilities: existing systems lack automatic recognition capabilities for three-dimensional scenes, and cannot automatically determine the type and requirements of the current design task based on user-created scenes and behavior data, resulting in mismatch between recommended content and user actual needs.

[0006] 2. Insufficient personalized recommendation: existing systems usually rely on static template and component libraries, making it difficult to provide personalized recommendations based on individual user needs and preferences, affecting user design efficiency.

[0007] 3. Insufficient dynamic response and adjustment capabilities: existing systems lack the ability to dynamically adjust recommended content during user design, and cannot respond to user design changes and feedback in real time, resulting in lagging or inaccurate recommended content.

[0008] 4. Lack of professional explanation of recommendations: existing systems lack the ability to provide professional explanations of recommended content, making it difficult for users to understand the basis of the recommendations, affecting the confidence and accuracy of design decisions.

[0009] In view of the above problems, the application provides a virtual construction system based on GraphRAG, which realizes dynamic adjustment and professional explanation of recommended content by intelligent scene recognition and personalized recommendation combined with industry knowledge graph, and significantly improves the intelligent level of the system and user experience.

[0010] The information disclosed in this section is only intended to deepen the understanding of the overall background of the application, and should not be regarded as acknowledging or implying in any form that this information constitutes prior art known to those skilled in the art. SUMMARY

[0011] The application aims to provide an electronic sand table personalized recommendation system in a virtual construction system based on GraphRAG, which realizes intelligent analysis of three-dimensional scene and user behavior data through the cooperative work of multiple functional modules, provides personalized sand table template and component recommendation, and supports dynamic update and explanation generation.

[0012] In order to achieve the above purpose, the application adopts the following technical solutions:

[0013] The application provides an electronic sand table personalized recommendation system in a virtual construction system based on GraphRAG, which comprises:

[0014] Data acquisition module: used for collecting user behavior data, scene perception data and industry knowledge data, and sending the collected data to the data processing and analysis module;

[0015] Data processing and analysis module: used for analyzing user behavior data and scene recognition and comparison; and sending the processing and analysis results to the GraphRAG retrieval enhancement module;

[0016] GraphRAG retrieval enhancement module: used for knowledge graph construction, template and component library retrieval and enhanced retrieval capability; and sending the processed data to the personalized recommendation module;

[0017] Personalized recommendation module: used for sand table template and sand table component recommendation, and sending the recommendation results to the personalized explanation generation module;

[0018] Personalized explanation generation module: using GraphRAG technology, combined with industry knowledge graph, generates personalized explanation for the recommended template and component, helps users understand the basis of the recommendation, and enhances user decision confidence; and sends the generated results to the visual interaction module;

[0019] Visual interaction module: including template, component interaction and dialogue interaction; and feeding back the interaction content to the dynamic adjustment and update module;

[0020] Dynamic adjustment and update module: used for scene change recognition and user feedback analysis, and feeds back the results of scene change recognition and user feedback analysis to the GraphRAG retrieval enhancement module.

[0021] Further, the data collection module specifically includes:

[0022] User behavior data collection submodule: this module is responsible for collecting user behavior data in the virtual construction system, including user operation records, design preferences, basic information, and historical design data; through analysis of these data, it provides basic support for subsequent scene recognition and recommendation;

[0023] Scene perception data collection submodule: this module collects data on the three-dimensional scenes established by users in the system, including terrain, building models, and environmental parameters in the scene; the system identifies the current design scene type through analysis of these data;

[0024] Industry knowledge data collection submodule: collects relevant data from the transportation infrastructure and urban planning industry knowledge base, builds an industry knowledge graph, and supports subsequent recommendation and explanation generation.

[0025] Further, the data processing and analysis module specifically includes:

[0026] User behavior data analysis: analyzes user behavior data through machine learning algorithms to identify user design preferences and operation habits, providing support for personalized recommendation;

[0027] Scene recognition and comparison: uses pattern recognition technology to analyze the three-dimensional scenes created by users and compares them with the pre-established scene model library to identify the current design scene type, providing a basis for recommendation decisions.

[0028] Further, the GraphRAG retrieval enhancement module specifically includes:

[0029] Knowledge graph construction: first, a comprehensive knowledge graph containing scenes, templates, components, and industry knowledge is constructed; through classification of virtual construction scenes, different types of scenes are defined, and these scenes are associated with corresponding sand table templates and components; at the same time, professional knowledge in the transportation infrastructure industry is integrated to establish semantic relationships between scenes, templates, components, and industry knowledge, ensuring the structured and semantically rich nature of the knowledge graph;

[0030] Template and component library retrieval: This module retrieves sand table templates and components that match the current scenario from the template and component database based on the GraphRAG technology. By combining the user's design requirements and scenario characteristics, GraphRAG can quickly locate relevant templates and components in the knowledge graph and provide accurate recommendations to ensure that the recommended content meets the user's design intent and scenario requirements.

[0031] Enhanced retrieval capability: Through semantic enhancement of the industry knowledge graph, the accuracy and relevance of retrieval are improved. GraphRAG uses semantic relationships in the industry knowledge graph to better understand the user's intent and requirements, ensuring that the recommended content not only matches the user's design scenario but also meets professional standards and industry best practices, further optimizing the user's design decision-making process.

[0032] Furthermore, the personalized recommendation module specifically includes:

[0033] Sand table template recommendation: The system analyzes the user's design requirements, behavior data, and current scenario characteristics, and uses the GraphRAG technology to retrieve sand table templates that match the current scenario in the knowledge graph, providing multiple alternative options for the user.

[0034] Sand table component recommendation: After the user selects a sand table template, the system further analyzes the characteristics of the template and the user's requirements and recommends matching components. The component recommendation process is not only based on the association information of components and templates in the knowledge graph but also combines the user's historical selection and feedback data. The system provides the best combination of components through intelligent optimization algorithms to ensure the coordination and functionality of components in the design.

[0035] Furthermore, the visual interaction module specifically includes:

[0036] Template and component interaction: Provides three-dimensional visual display of sand table components. Users can select, place, and adjust components through the interactive interface to achieve intuitive operation and real-time feedback of the design.

[0037] Dialogue interaction: Supports natural language dialogue interaction. Users can communicate with the system through voice or text instructions to obtain recommendations, modify design schemes, or obtain professional explanations.

[0038] Furthermore, the dynamic adjustment and update module specifically includes:

[0039] Scenario change recognition: The system monitors the user's design process in real time, recognizes scenario changes, and dynamically adjusts the recommended content to ensure real-time and accuracy of recommendations.

[0040] User feedback analysis: Based on user feedback, the recommendation algorithm is adjusted in real time to optimize recommended content, enabling the system to better adapt to changes in user needs and improve the intelligence and efficiency of the design.

[0041] By adopting the above technical solution, the present invention has the following beneficial effects:

[0042] This invention provides an intelligent, personalized, and dynamic design recommendation method for virtual construction systems, which can significantly improve design efficiency and user experience, and has broad application prospects and commercial value in fields such as transportation infrastructure and urban planning. Attached Figure Description

[0043] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0044] Figure 1 The diagram shows the structure of the personalized recommendation system for the electronic sand table in the GraphRAG-based virtual construction system provided in this embodiment of the invention. Detailed Implementation

[0045] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0046] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0047] Combination Figure 1 As shown, this embodiment provides a GraphRAG-based virtual construction system. The system works collaboratively through multiple functional modules to achieve intelligent analysis of 3D scene and user behavior data, provide personalized sandbox templates and component recommendations, and support dynamic updates and interpretation generation.

[0048] The specific components of this system include:

[0049] 1. Data Acquisition Module:

[0050] The user behavior data collection submodule is responsible for collecting user behavior data within the virtual construction system, including user operation records, design preferences, basic information (such as industry background and project type), and historical design data. Analysis of this data provides foundational support for subsequent scene recognition and recommendations.

[0051] Scene-aware data acquisition submodule: This module collects data from the 3D scene created by the user in the system, including terrain, building models, environmental parameters, etc. The system analyzes this data to identify the current design scene type.

[0052] Industry knowledge data collection submodule: Collects relevant data from industry knowledge bases such as transportation infrastructure and urban planning, constructs industry knowledge graphs, and supports subsequent recommendation and interpretation generation.

[0053] 2. Data Processing and Analysis Module:

[0054] User behavior data analysis: By analyzing user behavior data through machine learning algorithms, we can identify users' design preferences and operating habits, and provide support for personalized recommendations.

[0055] In this embodiment, users frequently select and adjust templates and components during the design process. The system needs to analyze user behavior to identify their preferences and then personalize recommendations accordingly. The specific implementation process is as follows:

[0056] 1) Deep learning algorithm selection: Preferably, this application adopts a CNN-LSTM combined model:

[0057] To analyze user behavior data, this application employs a combined model of convolutional neural networks (CNN) and long short-term memory networks (LSTM). This model can simultaneously capture the spatial and temporal features of user behavior, thereby more accurately identifying users' design preferences and operating habits.

[0058] 2) The specific algorithm flow is as follows:

[0059] 1. Data Acquisition and Preprocessing

[0060] (1) Data Input: The system collects various data types from user behavior, including click behavior, browsing history, selection frequency, operation sequence, etc. This behavioral data is encoded into time series data and input into the deep learning model.

[0061] (2) Data Feature Extraction: The system encodes user actions, converting each action (such as clicking, selecting, or adjusting templates and components) into a feature vector, forming a user behavior feature matrix. A specific example is as follows:

[0062] User click frequency: The number of times a user selects a certain template or component;

[0063] Operation sequence: Users add components in a fixed order (e.g., select a template first, then select components);

[0064] Time distribution: High-frequency user actions within a certain time period (such as frequently selecting a certain template in the later stages of a project).

[0065] 2. CNN Module: Spatial Feature Extraction

[0066] (1) Convolutional Layer: First, user behavior data undergoes spatial feature extraction through convolutional layers. Convolutional kernels are used to scan the user action behavior feature matrix to capture local features of user preferences, such as:

[0067] If a user repeatedly selects a particular template in multiple design projects, it indicates that the template may be a frequently used design element for the user.

[0068] When a user clicks on a certain type of component repeatedly within a short period of time, it indicates that the user has a high preference for that type of component.

[0069] (2) Pooling layer: Max Pooling layer is used to reduce the dimensionality of features, retain the most significant feature information, and enhance the robustness of features.

[0070] 3. LSTM module: Time series feature extraction

[0071] (1) Time series processing: The LSTM layer is input from the spatial features of the CNN to further analyze the temporal patterns of user behavior. For example, when users tend to choose the basic template in the early stages of a project, and prefer to add decorative components in the later stages, the LSTM can capture these changes in preferences over time.

[0072] (2) Hidden state update: LSTM continuously updates the hidden state through the memory and forget gate mechanism, records the user's behavioral habits and their evolution process, so as to better capture the user's design preferences at different stages.

[0073] 4. Fully Connected Layer and Output Layer

[0074] (1) The output of the LSTM is processed by a fully connected layer to generate a probability distribution of user preferences. The output layer assigns a recommendation probability to each electronic sandbox template and component to form a personalized recommendation list.

[0075] (2) Cross-Entropy Loss: Used to evaluate the model's prediction performance and to optimize the model during training using gradient descent.

[0076]

[0077] Among them, y i This indicates the actual user choice (1 for selected, 0 for not recommended); This represents the probability of the model making a recommendation.

[0078] 3) Specific examples of personalized recommendations

[0079] 1. Case Background

[0080] Suppose a user uses the system to design a sand table model of an intelligent transportation hub in a building project. The user exhibits the following behaviors during the design process:

[0081] Click and select the "Transportation Station" template multiple times;

[0082] The "high-density pedestrian flow" component is frequently selected to demonstrate the congestion situation at transportation hubs;

[0083] During the process, always prioritize selecting the basic template before adding functional components.

[0084] 2. Recommendation Process

[0085] (2) Data preprocessing

[0086] The system converts user behavior data (number of clicks, browsing time, selection order, etc.) into a behavior feature matrix and time series features.

[0087] Example of behavioral feature matrix: Users selected the "Transportation Station" template more frequently (eigenvalue = 0.9), while selecting the "Commercial Area" template less frequently (eigenvalue = 0.3).

[0088] (2) Feature extraction and preference analysis

[0089] CNN extracts user spatial preferences: Through convolutional layers, the system identifies that users prefer to select the "transportation station" template and click on the "high-density pedestrian flow" component multiple times, indicating that users pay more attention to the visualization of pedestrian flow when designing transportation hubs.

[0090] LSTM analysis of user time preferences: The LSTM layer found that users tend to choose basic templates in the early stages of a project, while they pay more attention to adding functional components in the later stages of the project. This indicates that users' design habits are to build the basic structure first and then add details.

[0091] 3. Generate recommendation results

[0092] (1) Based on the output of the CNN-LSTM model, the system generates a personalized electronic sand table template and component recommendation list, including:

[0093] Recommended template: Transportation station template, with a recommendation probability of 85%;

[0094] Recommended component: High-density pedestrian flow component, recommended probability is 90%;

[0095] Recommended order: First select the basic template, then add functional components.

[0096] (2) The recommendation results will be presented in an intuitive way on the user interface, along with the reasons for the recommendation (such as "You have selected the transportation station template multiple times in the last 5 projects").

[0097] Scene recognition and comparison: Using pattern recognition technology, the system analyzes the 3D scenes created by users and compares them with a pre-established scene model library to identify the current design scene type (such as bridge construction, road planning, etc.) and provide a basis for recommendation decisions.

[0098] 3. GraphRAG search enhancement module:

[0099] Knowledge Graph Construction: First, a comprehensive knowledge graph is constructed, encompassing scenarios, templates, components, and industry knowledge. By categorizing virtual construction scenarios, different types of scenarios are defined (e.g., bridge construction, waterway planning), and these scenarios are associated with corresponding sandbox templates and components. Simultaneously, professional knowledge from industries such as transportation infrastructure is integrated to establish semantic relationships between scenarios, templates, components, and industry knowledge, ensuring the structured nature and semantic richness of the knowledge graph.

[0100] In this embodiment, the knowledge graph is constructed in a modular and scalable manner, as detailed below:

[0101] I. Modular Design Form:

[0102] 1. Scene Module:

[0103] This module is responsible for constructing the structure of virtual construction scenes and classifying different types of scenes, such as "transportation hub", "commercial area", and "industrial park".

[0104] Implementation: Each scenario is a node in the knowledge graph, and it also has relevant attribute information (such as scenario type, scale, application field, etc.).

[0105] Example: In a transportation hub scenario, the system will construct a "Transportation Hub" node and associate it with corresponding visualization templates (such as "Subway Station Template" and "Bus Hub Template"). This modular structure allows the system to dynamically expand new nodes and relationships based on different scenario types.

[0106] 2. Sand table template module:

[0107] This module is responsible for building node information related to the visual dashboard template, including template type, layout structure, and applicable scenarios. The template module and the scenario module are linked through semantic relationships.

[0108] Implementation: Template nodes include attributes such as template ID, template name, template type, and template layout, and establish an "applicable" relationship with the corresponding scene nodes.

[0109] Example: When a user selects the "Subway Station" scenario, the system retrieves the "Subway Station Template" node and associates it with the "Transportation Hub" scenario node. This scalability of association allows new templates to be added to the knowledge graph without affecting the structure of existing templates.

[0110] 3. Sand table component module:

[0111] This module is responsible for building the node information of the visual dashboard components, including component type, functional attributes, and applicable templates. The component module and the template module are linked through semantic relationships.

[0112] Implementation: The component node contains attributes such as component ID, component name, component type, and function description, and establishes an "inclusion" relationship with the corresponding template node.

[0113] Example: In the "Metro Station Template," the system will construct nodes such as "Passenger Flow Component" and "Train Arrival Information Component," and associate them with template nodes through "containment" relationships. This modular design allows the system to dynamically adjust the graph structure according to changes in different components.

[0114] 4. Industry Knowledge Module:

[0115] This module integrates expertise in the transportation infrastructure field, establishing semantic relationships between scenarios, templates, components, and industry knowledge. For example, the "transportation hub" scenario may be associated with knowledge such as industry standards, design guidelines, and safety specifications.

[0116] Implementation method: Industry knowledge nodes include industry standards, design specifications, best practices, etc. The system connects these nodes with related scenarios, templates and components through semantic relationships (such as "complies with" and "based on").

[0117] Example: In the "Transportation Hub" scenario node, the system can associate an "Industry Standard" node, which contains standards such as "Pedestrian Flow Control" and "Evacuation Guidance," and establish "compliance" relationships with relevant templates and components. This modular structure ensures the dynamic updating and expansion of industry knowledge.

[0118] II. Scalable Design Forms:

[0119] 1. Dynamic addition of new scenes and templates:

[0120] The system supports the dynamic addition of new scenes and templates through API interfaces. For example, when the system needs to add a new "smart parking lot" scene, it only needs to add a new node and connect it to the corresponding template and components through the "applicable" relationship.

[0121] Extension method: Adding new scenes and templates will not affect the existing graph structure because the system adopts a modular node design, which allows new nodes to be connected in parallel or cross-connected with existing nodes.

[0122] 2. Dynamic updating of semantic relationships:

[0123] When the system receives new industry standards or design specifications, the industry knowledge module can automatically update the corresponding nodes and relationships. For example, when a new industry standard for "pedestrian-vehicle separation" is added, the node can automatically establish a "compliance" relationship with the "transportation hub" scenario and the corresponding templates and components.

[0124] Implementation: Through GrgphRAG technology, the system can automatically adjust and optimize the semantic structure of the graph based on the contextual relationships of new knowledge.

[0125] 3. Cross-industry knowledge integration:

[0126] The system supports the dynamic integration of cross-industry knowledge, such as extending knowledge from intelligent transportation to the field of intelligent buildings. This expansion capability relies on the structure of a modular knowledge graph, allowing the system to add new industry knowledge nodes and their associations without changing the basic structure.

[0127] Implementation: Through modular industry knowledge node design, the system can easily add new domain knowledge nodes and make cross-domain connections through semantic relationships such as "applicable" or "based on".

[0128] III. Example: Modular and scalable knowledge graph construction

[0129] Suppose the system needs to build a knowledge graph for an intelligent transportation project. The specific implementation process is as follows:

[0130] 1. Scene node construction:

[0131] The system first adds a "Smart Transportation Hub" scenario node, defines the scenario type as "Transportation Infrastructure", the scale as "Large", and the application field as "Urban Transportation".

[0132] 2. Template node construction:

[0133] In the "Intelligent Transportation Hub" scenario, the system adds a "Public Transportation Hub Template" node and establishes an "Applicable" relationship with the scenario nodes.

[0134] The "Public Transport Hub Template" node has attributes such as template type "Transportation Station" and layout structure "Multi-level Streamline Layout".

[0135] 3. Component node construction:

[0136] The system adds "Pedestrian Flow Monitoring Component" and "Vehicle Dispatch Component" nodes under the "Public Transportation Hub Template" and establishes an "inclusion" relationship with the template node.

[0137] The component node includes the component type "Data Visualization", the function description "Real-time Pedestrian Flow Monitoring and Scheduling", and the applicable template is "Transportation Station Template".

[0138] 4. Industry knowledge node integration:

[0139] The system adds a "Pedestrian and Vehicle Separation Industry Standard" node to the "Intelligent Transportation Hub" scenario node and associates it with scenarios, templates, and components through the "compliance" relationship.

[0140] The "Pedestrian and Vehicle Separation Industry Standard" node has a standard type of "Traffic Design Standard" and an applicable scope of "High-Density Pedestrian Areas". The system adjusts the design suggestions of the scenario and template according to the standard requirements.

[0141] Through modular and scalable design, the system's knowledge graph can dynamically adapt to different virtual construction scenarios, visualization dashboard templates and components, as well as changes in industry knowledge in the transportation infrastructure field. This design not only improves the system's flexibility and scalability but also ensures the structured and semantically rich nature of the knowledge graph.

[0142] Template and Component Retrieval: This module, based on GraphRAG technology, retrieves sandbox templates and components matching the current scenario from the template and component database. By combining user design requirements and scenario characteristics, GraphRAG can quickly locate relevant templates and components in the knowledge graph and provide accurate recommendations, ensuring that the recommended content matches the user's design intent and scenario requirements.

[0143] Enhanced Search Capabilities: GraphRAG improves search accuracy and relevance through semantic enhancement of industry knowledge graphs. Leveraging semantic relationships within industry knowledge graphs, GraphRAG better understands user intent and needs, ensuring that recommended content not only highly matches the user's design scenario but also meets professional standards and industry best practices, further optimizing the user's design decision-making process.

[0144] 4. Personalized Recommendation Module:

[0145] Sandbox Template Recommendation: By analyzing users' design needs, behavioral data, and current scene characteristics, the system uses GraphRAG technology to retrieve sandbox templates matching the current scene from a knowledge graph, providing users with multiple alternatives. These options may cover different design styles and component types, allowing users to choose according to specific project needs and personal preferences, ensuring design diversity and flexibility.

[0146] Sandbox Component Recommendation: After a user selects a sandbox template, the system further analyzes the template's characteristics and the user's needs, recommending matching components. This recommendation process is based not only on the association information between components and templates in a knowledge graph but also on the user's historical choices and feedback data. Through intelligent optimization algorithms, the system provides the optimal combination of components to ensure their coordination and functionality within the design. This optimized combination not only improves overall design efficiency but also guarantees the accuracy and professionalism of the design results, enabling users to complete high-quality design solutions in a shorter time.

[0147] In this embodiment, to further illustrate how this application provides the optimal combination of components through intelligent optimization algorithms, a detailed explanation is provided below with specific examples:

[0148] Case Background: Suppose a user is designing a large-screen visualization sandbox for a city traffic dispatch center. The system needs to recommend the most suitable component combination based on the user's selected visualization template. These components include: real-time traffic flow map, traffic accident alarm window, dispatch strategy chart, and weather data display, etc.

[0149] User needs analysis and sandbox template recommendations:

[0150] (1) Requirements Analysis: The system first analyzes the user's design requirements and the characteristics of the current scenario. For example, in the scenario of a traffic dispatch center, the user's main focus is on real-time traffic flow, accident alarm information, and dispatch strategy charts.

[0151] (2) Template Recommendation: The system uses GraphRAG technology to retrieve sandbox templates (such as "City Real-time Traffic Template") that match "Traffic Dispatch Center" from the knowledge graph and recommends these templates to the user. These templates have layouts and basic component structures suitable for traffic management scenarios.

[0152] Recommendation and intelligent optimization algorithm for electronic sand table components:

[0153] (1) After the user selects the “City Real-time Traffic Template”, the system further analyzes the characteristics of the template (such as focusing on displaying real-time data and traffic status information) and the user’s specific needs, and recommends a matching visualization screen component.

[0154] (2) Recommendation Algorithm Selection: Multi-objective Optimization Algorithm:

[0155] The system uses a multi-objective optimization algorithm to recommend the best combination of components to users, taking into account the functionality, coordination, user preferences, and layout characteristics of the components.

[0156] Multi-objective optimization algorithm description:

[0157] (1) Objective function: The objective function of the multi-objective optimization algorithm is as follows:

[0158] O total =max(α·P) u +β·F c +γ·C t );

[0159] Among them, O total P represents the total optimization score for the component combination. u User preference, i.e., the frequency of occurrence of components selected by the user in the past; F c For functional scoring, the degree of functional matching between components and templates is evaluated; C t The score represents the coordination between components, i.e., the layout compatibility of different components under the same template; α, β, and γ are the weights of each objective, which are adjusted according to actual needs.

[0160] Recommendation process for electronic sand table component combinations:

[0161] (1) Step 1: Input user preferences and template features

[0162] User preference data and selected template features are used as input data. For example, users prefer "real-time traffic flow map" and "accident alarm window", while the template focuses on displaying traffic status information and dynamic scheduling strategies.

[0163] (2) Step 2: Calculate the component functionality score

[0164] The system retrieves association information between components and templates using a knowledge graph. For example, the functionality score for "Real-time Traffic Flow Map" and "City Real-time Traffic Template" is 85%, while the functionality score for "Accident Alarm Window" is 90%.

[0165] (3) Step 3: Assess the coordination between components

[0166] The system evaluates the layout compatibility of different components within the same template based on historical combination data of components in the knowledge graph. For example, the coordination score between "Real-time Traffic Flow Map" and "Dispatch Strategy Chart" is 80%, and the coordination score between "Traffic Accident Alarm Window" and "Weather Data Display" is 70%.

[0167] Intelligent optimization and selection of optimal component combinations:

[0168] (1) Calculate the total optimization score: The system uses a multi-objective optimization algorithm to calculate the optimization score for different component combinations. Assume there are four possible component combinations:

[0169] Combination A: Includes "Real-time Traffic Flow Map" and "Dispatch Strategy Chart", with a total score of 0.85;

[0170] Combination B: Includes "Traffic Accident Reporting Window" and "Weather Data Display", with a total score of 0.80;

[0171] Combination C: Includes "Real-time Traffic Flow Map", "Traffic Accident Alarm Window" and "Dispatch Strategy Chart", with a total score of 0.90;

[0172] Combination D: Includes "Traffic Accident Alarm Window", "Meteorological Data Display" and "Dispatch Strategy Chart", with a total score of 0.75.

[0173] (2) The system selects the highest-scoring combination (such as combination C) and recommends it to the user.

[0174] Visualization of recommendation results and user feedback:

[0175] (1) Visualization: The system displays recommended component combinations on the user interface and provides reasons for the recommendations. For example, "Combination C scores the highest because it contains the main components that users prefer, and it has the best functionality and coordination."

[0176] (2) User feedback and dynamic adjustments:

[0177] If a user adjusts the recommendations (such as changing components or rearranging the layout), the system will recalculate the optimization score based on the new operation data and update the recommended combination.

[0178] Algorithm optimization and continuous learning:

[0179] (1) Real-time feedback mechanism: The system adjusts the weight parameters of the multi-objective optimization algorithm based on user selections and feedback to better adapt to user needs. For example, if a user selects "real-time traffic flow map" multiple times, the system will increase its weight in future recommendations.

[0180] (2) Model update: The system adds new user selection records to the knowledge graph through GraphRAG technology to improve the intelligence and personalization of the recommendation results.

[0181] Through the specific embodiments described above, the system utilizes GraphRAG technology and a multi-objective optimization algorithm to recommend the optimal component combination that matches the visualization dashboard template for users. The optimization algorithm comprehensively calculates the optimal component recommendation combination through multi-objective calculations considering user preferences, component functionality, and coordination. Simultaneously, the system features real-time feedback and dynamic adjustment capabilities to ensure the accuracy of the recommendation results and user satisfaction.

[0182] 5. Personalized Explanation Generation Module:

[0183] By leveraging GraphRAG technology and combining it with industry knowledge graphs, personalized explanations are generated for recommended templates and components, helping users understand the basis for the recommendations and enhancing their confidence in making decisions. Through generative models, professional knowledge is transformed into easily understandable explanatory text, improving users' comprehension and acceptance of the recommended content.

[0184] 6. Visual Interaction Module:

[0185] Template and component interaction: Provides a 3D visualization of sandbox components. Users can select, place, and adjust components through the interactive interface, achieving intuitive operation and real-time feedback in design.

[0186] Dialogue Interaction: Supports natural language dialogue interaction, allowing users to communicate with the system via voice or text commands to obtain recommendations, modify design solutions, or get professional explanations.

[0187] 7. Dynamically adjust and update module:

[0188] Scene change recognition: The system monitors the user's design process in real time, identifies scene changes (such as bridge construction changing to highway construction), and dynamically adjusts the recommended content to ensure the real-time nature and accuracy of the recommendations.

[0189] In this embodiment, to further illustrate how this application dynamically adjusts the recommended content, a detailed explanation is provided below with specific examples:

[0190] Preferably, this application adds a reinforcement learning algorithm to the dynamic adjustment module.

[0191] Example: 3D scene change recognition and dynamic recommendation adjustment:

[0192] Case Background: Suppose a user is designing a series of 3D scenes. The initial scene is a "transportation construction scene," containing models of roads, bridges, and traffic lights. The user then switches to a "water conservancy construction scene," containing models of dams, canals, and hydrological monitoring stations. The system needs to recognize changes in the 3D scenes in real time and adjust the recommended sand table components (visualized large screen components).

[0193] 1. Application of the scene-aware data acquisition submodule:

[0194] (1) Data acquisition process:

[0195] The system monitors the user's 3D design scene in real time through the scene-aware data acquisition submodule, and the information collected includes:

[0196] Topographic data: such as flat roads, rolling hills, rivers and dams, etc.;

[0197] Architectural models: such as bridges, tunnels, canals, and hydrological monitoring stations;

[0198] Environmental parameters: such as traffic flow, rainfall, wind speed, etc.

[0199] (2) Data analysis and scene recognition:

[0200] The system uses this data to analyze the current 3D scene type. For example, when the system identifies the terrain data as rivers and valleys and detects the building models as dams and hydrological monitoring stations, it marks the scene type as "water conservancy construction scene".

[0201] Changes in the identified scenarios will trigger corresponding adjustments to the recommendation strategy. For example, when the system changes from a "transportation construction scenario" to a "water conservancy construction scenario," it will prioritize recommending components related to water conservancy construction (such as channel monitoring charts and dam control models).

[0202] 2. Dynamic adjustment and optimization of reinforcement learning algorithms:

[0203] To ensure the real-time nature and accuracy of the recommendations, the system uses reinforcement learning algorithms to dynamically adjust and optimize after recognizing scene changes.

[0204] (1) Reinforcement learning model structure

[0205] Reinforcement learning model selection: Deep Q-Network (DQN):

[0206] DQN is a reinforcement learning algorithm suitable for dynamic scene changes. It uses neural networks to estimate the Q-values ​​of states and actions to learn and adjust the optimal action strategy.

[0207] Definitions of state, action, and reward:

[0208] State (S): The current 3D scene type (such as a transportation construction scene or a water conservancy construction scene), the selected components, and the user's design preferences.

[0209] Action (A): The component adjustment action recommended by the system in the current state, such as "recommend hydrological monitoring charts" or "replace road surveillance cameras".

[0210] Reward (R): The effectiveness of actions is evaluated based on user feedback. When users accept recommended components or adjust actions, the system gives a positive reward; when users reject recommendations, a negative reward is given.

[0211] Reward function: R = α·U s +β·M c -γ·D t ;

[0212] Among them, U s Indicates user satisfaction with the recommended content (e.g., click-through rate or selection rate); M c This indicates the degree of matching between the recommended component and the current scene; D t This represents the recommended delay time; α, β, and γ are the weights of each parameter.

[0213] (2) The training and dynamic adjustment process of reinforcement learning

[0214] Q-value update and strategy optimization:

[0215] After each scene change, the system updates the Q-value based on the new state and the action taken, optimizing the next action. The formula for updating the Q-value is:

[0216]

[0217] Where Q(s, a) represents the Q-value of action a under condition s; η represents the learning rate, indicating the speed at which the Q-value is updated; R represents the current reward value; γ represents the discount factor, measuring the importance of future rewards; max a′ Q(s′,a′) represents the maximum possible Q value under the new state s′.

[0218] (3) Specific applications of reinforcement learning in 3D scene changes

[0219] Scene change recognition and action selection:

[0220] When the system recognizes that the scenario has switched from "transportation construction" to "water conservancy construction", the DQN model calculates the optimal recommended action based on the new state, such as "recommending hydrological monitoring charts".

[0221] Real-time recommendations and feedback:

[0222] The system will display recommended components to the user. If the user selects the recommended "hydrological monitoring chart", the system will give a positive reward and update the Q value; if the user selects other components, the system will give a negative reward and adjust the strategy.

[0223] Continuous optimization and dynamic adjustment:

[0224] The reinforcement learning algorithm updates itself after each scene change and user feedback, ensuring real-time optimization of the recommendation strategy. For example, if a user selects the dam control component multiple times in the "water conservancy construction scenario," the system will increase its priority in future recommendations.

[0225] 3. Advantages and effects:

[0226] (1) Real-time response: The reinforcement learning algorithm can adjust the strategy according to the real-time changes of the three-dimensional scene to ensure a high degree of matching between the recommended content and the scene.

[0227] (2) Continuous learning: The system continuously learns from user feedback and changes in scenarios to optimize recommendation strategies, improve the accuracy of recommended content and user satisfaction.

[0228] This embodiment combines a scene-aware data acquisition submodule with a reinforcement learning algorithm to achieve real-time recognition of 3D scene changes and dynamic optimization of recommendation strategies. The reinforcement learning algorithm enables the system to quickly adjust recommended content after scene switching, improving the real-time performance and accuracy of recommendations, thereby enhancing the system's intelligence and practicality in virtual construction.

[0229] User feedback analysis: Based on user feedback, the recommendation algorithm is adjusted in real time to optimize recommended content, enabling the system to better adapt to changes in user needs and improve the intelligence and efficiency of the design.

[0230] Compared with the prior art, the present invention has the following significant technical advantages and beneficial effects:

[0231] 1. Intelligent scene perception:

[0232] By comprehensively analyzing 3D scene and user behavior data, the system achieves intelligent perception and recognition of complex design scenarios. It can provide highly matched recommended content based on the characteristics of different scenarios, significantly improving the intelligence level of the virtual construction system.

[0233] 2. Personalized recommendations are more precise:

[0234] By leveraging GraphRAG technology combined with industry knowledge graphs, the system can provide accurate personalized recommendations based on users' individual needs and design preferences, improving design efficiency and helping users quickly find the most suitable sandbox templates and components.

[0235] 3. Dynamic adjustment in real time:

[0236] The system supports real-time analysis and dynamic adjustment of scenario changes and feedback during the user's design process, ensuring that recommended content always remains consistent with the user's design needs, thereby improving the responsiveness and intelligence of the design process.

[0237] 4. Professional interpretation of personalized services:

[0238] By combining industry knowledge graphs, the system generates professional and personalized explanations for recommended content, helping users understand the basis for recommendations, enhancing users' confidence in design, and improving the user experience of the system.

[0239] 5. Visualization and user-friendliness:

[0240] The system provides an intuitive 3D visualization interface and natural language dialogue interaction, making the user operation more convenient and the feedback more timely during the design process, thus improving the overall user experience and design efficiency.

[0241] This invention provides an intelligent, personalized, and dynamic design recommendation method for virtual construction systems, which can significantly improve design efficiency and user experience, and has broad application prospects and commercial value in fields such as transportation infrastructure and urban planning.

[0242] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. An electronic sand table personalized recommendation system in a GraphRAG-based virtual construction system, characterized in that, Comprise: Data collection module: for collecting user behavior data, scene perception data and industry knowledge data, and sending the collected data to the data processing and analysis module; The data processing and analysis module: for analyzing user behavior data and scene recognition and comparison; and sending the processing and analysis results to the GraphRAG retrieval enhancement module; The GraphRAG retrieval enhancement module: for knowledge graph construction, template and component library retrieval and enhanced retrieval capability; and sending the processed data to the personalized recommendation module; The GraphRAG retrieval enhancement module specifically includes: Knowledge graph construction: first, a comprehensive knowledge graph containing scenes, templates, components and industry knowledge is constructed; by classifying virtual construction scenes, different types of scenes are defined, and these scenes are associated with corresponding sand table templates and components; at the same time, the professional knowledge of the transportation infrastructure industry is integrated, the semantic relationship between scenes, templates, components and industry knowledge is established, and the structure and semantic richness of the knowledge graph are ensured; Template and component library retrieval: this module is based on GraphRAG technology, and retrieves sand table templates and components matching the current scene in the template and component database; by combining the user's design requirements and scene characteristics, GraphRAG can quickly locate related templates and components in the knowledge graph, and provide accurate recommendations to ensure that the recommended content meets the user's design intent and scene requirements; Enhanced retrieval capability: through semantic enhancement of industry knowledge graph, the accuracy and relevance of retrieval are improved; GraphRAG uses the semantic relationship in the industry knowledge graph to better understand the user's intent and requirements, ensuring that the recommended content not only matches the user's design scene highly, but also meets professional standards and industry best practices, further optimizing the user's design decision-making process; The personalized recommendation module: for sand table template and sand table component recommendation, and sending the recommendation results to the personalized explanation generation module; the personalized recommendation module specifically includes: Sand table template recommendation: the system analyzes the user's design requirements, behavior data and current scene characteristics, and uses GraphRAG technology to retrieve sand table templates matching the current scene in the knowledge graph, providing multiple alternative solutions for the user; Sand table component recommendation: after the user selects a sand table template, the system further analyzes the characteristics of the template and the user's requirements, and recommends matching components; the component recommendation process is not only based on the association information of components and templates in the knowledge graph, but also combines the user's historical selection and feedback data; the system provides the best combination of components through intelligent optimization algorithm to ensure the coordination and functionality of components in design; The personalized explanation generation module: uses GraphRAG technology to generate personalized explanations for recommended templates and components in combination with industry knowledge graph, helping users understand the basis of recommendations and enhancing user decision-making confidence; and sending the generated results to the visualization interaction module; The visualization interaction module: includes template, component interaction and dialogue interaction; and feedbacks the interaction content to the dynamic adjustment and update module; The dynamic adjustment and update module is used for scene change identification and user feedback analysis, and feeds back the scene change identification and user feedback analysis results to the GraphRAG retrieval enhancement module.

2. The GraphRAG-based virtual construction system electronic sand table personalization recommendation system according to claim 1, characterized in that, The data collection module specifically includes: User behavior data collection submodule: This module is responsible for collecting user behavior data in the virtual construction system, including user operation records, design preferences, basic information, and historical design data; through analysis of these data, it provides basic support for subsequent scene identification and recommendation; Scene perception data collection submodule: This module collects data on the three-dimensional scenes created by users in the system, including terrain, building models, and environmental parameters in the scene; through analysis of these data, the system identifies the current design scene type; Industry knowledge data collection submodule: Collects relevant data from the transportation infrastructure and urban planning industry knowledge base to build an industry knowledge graph, supporting subsequent recommendation and explanation generation.

3. The GraphRAG-based virtual construction system electronic sand table personalized recommendation system according to claim 1, characterized in that, The data processing and analysis module specifically includes: User behavior data analysis: Analyze user behavior data through machine learning algorithms to identify user design preferences and operation habits, providing support for personalized recommendations; Scene identification and comparison: Use pattern recognition techniques to analyze the three-dimensional scenes created by users and compare them with the pre-established scene model library to identify the current design scene type, providing a basis for recommendation decisions.

4. The GraphRAG-based virtual construction system electronic sand table personalization recommendation system according to claim 1, characterized in that, The visual interaction module specifically includes: Template and component interaction: Provides three-dimensional visual display of sand table components; users select, place, and adjust components through the interactive interface to achieve intuitive operation and real-time feedback of design; Dialogue interaction: Supports natural language dialogue interaction; users communicate with the system through voice or text instructions to obtain recommendations, modify design schemes, or obtain professional explanations.

5. The GraphRAG-based virtual construction system electronic sand table personalization recommendation system according to claim 1, characterized in that, The dynamic adjustment and update module specifically includes: Scene change identification: The system monitors the user's design process in real time, identifies scene changes, and dynamically adjusts the recommended content to ensure real-time and accuracy of the recommendations; User feedback analysis: Adjusts the recommendation algorithm in real time according to user feedback information to optimize the recommended content, enabling the system to better adapt to changes in user needs and improve design intelligence and efficiency.

Images