Three-dimensional twin scene construction method and device

Abstract

The application discloses a three-dimensional twin scene construction method and device, and belongs to the technical field of digital twinning. The method comprises the following steps: constructing a semantic ontology model, mapping a physical asset to a semantic node containing business capability, real-time state and logical association; based on the semantic ontology model, in response to a spatial operation of a user on a three-dimensional digital twinning base, only when a physical asset pointed by the spatial operation has been mapped to a semantic node, a corresponding business intent label is captured, and the business intent label is converted into a structured query instruction; according to the structured query instruction, a corresponding three-dimensional model, a dynamic visual component and an interaction panel are automatically indexed and assembled from a predefined visual asset library; based on a camera perspective and an entity bounding box, a non-occluded arrangement position of the interaction panel in a three-dimensional space is calculated in real time. The application can solve the problems of non-automation of the process, logic hard coding, business semantic fault and visual interaction redundancy in the prior art.

Description

Technical Field

[0001] This application belongs to the field of digital twin technology, and specifically relates to a method and apparatus for constructing a three-dimensional twin scene. Background Technology

[0002] In the visual management of tower operations, the needs of different management departments for the visualization platform are highly fragmented. Existing digital twin platforms mostly adopt the full display or static template switching method. The former loads all resources at once, resulting in information overload, while the latter pre-develops templates for users to select and switch. It is difficult to meet the core need of users to quickly build a targeted 3D sand table based on specific business intentions within any selected area.

[0003] Currently, in the field of web-based 3D visualization scene configuration, the mainstream approach is a web-based configuration solution. Its core implementation logic is as follows: pre-develop 3D templates adapted to different business scenarios, establish a mapping relationship between database fields and template placeholders to complete the static binding of data and templates, and finally, users switch the corresponding template by selecting parameters through the backend management interface to update and configure the 3D scene.

[0004] However, this solution is essentially a static selection mode, which has the following limitations: the process is not automated and relies on manual orchestration; the logic is hard-coded, and code needs to be modified to adjust the rules; there is a gap in business semantics, making it impossible to accurately filter based on business intent; and the visual interaction is redundant, with UI panels severely overlapping and obscuring each other. Summary of the Invention

[0005] To address the aforementioned issues, this application provides a method and apparatus for constructing a 3D twin scene. By establishing an automated generation framework based on semantic driving, declarative rules, and dynamic assembly, it can solve problems such as non-automated processes, hard-coded logic, fragmented business semantics, and redundant visual interactions in existing technologies. The method includes: Construct a semantic ontology model to map physical assets into semantic nodes that contain business capabilities, real-time status, and logical relationships; Based on the semantic ontology model, in response to the user's spatial operations on the three-dimensional digital twin base, the corresponding business intent tag is captured only when the physical asset pointed to by the spatial operation has been mapped as a semantic node, and the business intent tag is converted into a structured query instruction. Based on structured query instructions, the system automatically indexes and assembles corresponding 3D models, dynamic visual components, and interactive panels from a predefined visual asset library; and calculates the unobstructed arrangement of the interactive panels in 3D space in real time based on camera view and entity bounding boxes.

[0006] In this embodiment of the application, a semantic ontology model is constructed to map physical assets into semantic nodes that include business capabilities, real-time status, and logical relationships, including: Abstract physical assets into ontology classes, and define static slots for describing their inherent attributes and dynamic slots for describing their functional characteristics for each ontology class. Extract the attribute data of specific physical assets from the underlying resource database, populate the static and dynamic slots of the corresponding ontology class, and generate semantic node instances that correspond one-to-one with the physical assets. Based on the spatial topological and business logic relationships between physical assets, object attribute associations are constructed between semantic node instances to form a network structure with business semantic associations.

[0007] In this embodiment, in response to a user's spatial operation on a 3D digital twin platform, the corresponding business intent tag is captured only when the physical asset pointed to by the spatial operation has been mapped as a semantic node, and the business intent tag is converted into a structured query instruction, including: In response to the user's spatial operation on the 3D digital twin base, determine whether the physical asset to which the spatial operation is directed has been mapped as a semantic node; if not, terminate the process; if so, extract the geofence information corresponding to the spatial operation and encapsulate the geofence information into a spatial query predicate. If the mapping exists, obtain the business intent tag specified by the user, perform semantic parsing on the business intent tag according to the preset business semantic dictionary, and map the business intent tag to at least one business capability predicate; The spatial query predicate and the business capability predicate are logically combined to generate a structured query instruction that conforms to the preset syntax, and the structured query instruction is sent to the declarative rule auto-assembly operator.

[0008] In this embodiment of the application, based on structured query instructions, the corresponding 3D model, dynamic visual components, and interactive panel are automatically indexed and assembled from a predefined visual asset library, including: Receive and parse structured query commands, extract business intent tags and entity type information from the commands, and index 3D model templates and visual component templates that match the business intent from the visualization asset library according to the preset business intent-visual component mapping table. Obtain the real-time status data of each entity in the entity set corresponding to the structured query command, match the real-time status data with the condition part of the preset condition-action rule set, and filter out the entities that need to trigger special visual performance. Based on the index and matching results, perform assembly operations, which include: instantiating a 3D model template to generate a 3D model instance and attaching it to the corresponding entity location; instantiating a visual component template to generate a dynamic visual component and associating it with the corresponding entity; and dynamically calling and binding the corresponding interactive panel for each entity.

[0009] In this embodiment of the application, the non-occluded arrangement position of the interactive panel in three-dimensional space is calculated in real time based on the camera view and the physical bounding box, including: Obtain the view matrix and projection matrix of the current camera view, project the 3D bounding box of each entity and the candidate position of each interactive panel onto the screen coordinate system, and generate the occupied area of ​​each entity in the screen space and the candidate area of ​​each interactive panel in the screen space. Calculate the degree of occlusion between the candidate area of ​​each interactive panel and the area occupied by the corresponding entity, as well as the degree of overlap between the candidate areas of each interactive panel. When the degree of occlusion or overlap exceeds a preset threshold, trigger the position adjustment mechanism. Based on a physical simulation-based repulsion model or a gradient descent-based optimization algorithm, the position offset vectors of each interactive panel in three-dimensional space are iteratively adjusted to minimize the degree of occlusion and overlap until the preset visual clarity constraints are met.

[0010] In this embodiment, the degree of occlusion between the candidate area of ​​each interactive panel and the area occupied by the corresponding entity, as well as the degree of overlap between the candidate areas of each interactive panel, are calculated. When the degree of occlusion or overlap exceeds a preset threshold, a position adjustment mechanism is triggered, including: Project the 3D bounding box of each entity onto the screen space to generate the polygon of the area occupied by the entity. Project the candidate position of each interactive panel onto the screen space to generate the polygon of the candidate area of ​​the panel. Calculate the intersection-union ratio between the candidate area of ​​the panel and the area occupied by the corresponding entity, and determine the intersection-union ratio as the occlusion degree value. Obtain candidate region polygons for all interactive panels within the same screen space, iterate through and calculate the intersection-union ratio between every two candidate regions of different interactive panels, and determine the maximum value or cumulative value of all intersection-union ratios as the degree of overlap. The occlusion level value is compared with the preset occlusion threshold, and the overlap level value is compared with the preset overlap threshold. If the occlusion level value exceeds the occlusion threshold or the overlap level value exceeds the overlap threshold, it is determined that there is a visual conflict, and the position adjustment mechanism is triggered.

[0011] In this embodiment, a repulsive force model based on physical simulation or an optimization algorithm based on gradient descent is used to iteratively adjust the position offset vector of each interactive panel in three-dimensional space, with the goal of minimizing the degree of occlusion and overlap, until a preset visual clarity constraint is met, including: A total energy function is constructed with the position offset vectors of all interactive panels as independent variables. The total energy function includes an occlusion penalty term, an overlap penalty term, and a position constraint term. The occlusion penalty term is used to quantify the degree of occlusion between the interactive panel and the corresponding entity, the overlap penalty term is used to quantify the degree of overlap between interactive panels, and the position constraint term is used to limit the distance of the interactive panel from its initial anchor position. The weight coefficients of the occlusion penalty term, overlap penalty term, and position constraint term in the total energy function can be adaptively adjusted according to the needs of the actual business scenario. This embodiment does not limit their fixed values. The algorithm employs either gradient descent-based optimization or physical simulation-based repulsion model to minimize the total energy function. It iteratively updates the position offset vectors of each interactive panel. In each iteration, the offset is adjusted based on the gradient direction of the current total energy function with respect to the offset vector, or based on the sum of the repulsion between panels and the attraction between the entity and the panel. During the iteration process, the damping coefficient can be set to suppress the oscillation of the position offset, further ensuring the convergence stability of the algorithm. After each iteration, the current total energy function value is recalculated. When the total energy function value is less than the preset energy threshold, or the change in the energy function value after multiple consecutive iterations is less than the preset change threshold, or the number of iterations reaches the preset maximum number of iterations, the iteration stops and the current position offset vector is output as the final arrangement result.

[0012] This application also provides a three-dimensional twin scene construction device, the device comprising: The ontology module is used to build a semantic ontology model, mapping physical assets into semantic nodes that contain business capabilities, real-time status, and logical relationships. The structure query module is used to respond to users' spatial operations on the 3D digital twin base based on the semantic ontology model. Only when the physical asset pointed to by the spatial operation has been mapped as a semantic node, it captures the corresponding business intent tag and converts the business intent tag into a structured query instruction. The automatic assembly module is used to automatically index and assemble corresponding 3D models, dynamic visual components and interactive panels from a predefined visual asset library according to structured query instructions; and calculates the non-occluded arrangement position of the interactive panels in 3D space in real time based on the camera view and the entity bounding box.

[0013] This application also provides a computer-readable storage medium having a computer program stored thereon, wherein when the program is executed by a processor, it implements the steps of the method provided in the above embodiments.

[0014] This application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the method provided in the above embodiments.

[0015] Other features and advantages of this application will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the application. The objectives and other advantages of this application may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating a method for constructing a three-dimensional twin scene, as provided in an embodiment of this application.

[0018] Figure 2 A block diagram of a three-dimensional twin scene construction device provided in an embodiment of this application.

[0019] Figure 3 This is a logical architecture diagram of a three-dimensional twin scene construction device provided in an embodiment of this application.

[0020] Figure 4 This is a schematic diagram illustrating the automatic generation of a three-dimensional twin scene, as provided in an embodiment of this application.

[0021] Figure 5 A comparison diagram showing the effect of an intelligent obstacle avoidance algorithm provided in an embodiment of this application.

[0022] Figure 6 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments 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.

[0024] Currently, in the visual management of tower operation, the needs of different management departments (such as emergency support, resource planning, and fire prevention monitoring) for visual bases are highly fragmented: some need to view the distribution of disaster-stricken tower sites, some are concerned about the availability of mounting resources, and some focus on fire prevention monitoring points.

[0025] Existing digital twin platforms mostly employ either full-scale display or static template switching. Full-scale display involves loading all resource entities into a 3D scene at once, leading to information overload and visual clutter. Static template switching involves pre-developing several 3D templates for different business scenarios, which users can switch between via parameters in the backend. Neither of these methods adequately meets the core needs of users in real-world work—namely, the ability to quickly construct a targeted and interactive 3D sandbox within any selected geographical area, based on a specific business intent (such as forest fire prevention).

[0026] Currently, in the field of web-based 3D visualization scene configuration, the mainstream approach is a web-based configuration solution. Its core implementation logic is as follows: pre-develop 3D templates adapted to different business scenarios, establish a mapping relationship between database fields and template placeholders to complete the static binding of data and templates, and finally, users switch the corresponding template by selecting parameters through the backend management interface to update and configure the 3D scene.

[0027] Limitations Analysis: This scheme is essentially a static selection pattern, and its limitations are mainly reflected in the following four aspects: (1) Non-automated process The creation of existing 3D sand tables relies heavily on manual arrangement. The placement of models and the mounting method of UI panels all require the intervention of graphic designers and front-end developers, making it impossible to achieve automated generation that can be used immediately.

[0028] (2) Logic hard coding The scene display is strongly tied to the business logic. Fine-tuning of business rules (such as changing filtering conditions or adjusting display styles) requires modifying the underlying code, resulting in long response cycles, poor flexibility, and difficulty in adapting to rapid changes in business needs.

[0029] (3) Business semantic gap The rendering engine can only understand three-dimensional spatial coordinate information and cannot recognize the business attributes of entities such as towers and data centers (such as the type of mounted equipment, backup power duration, and power supply level). When users request to view emergency support resources, the system cannot perform accurate resource filtering and logical deduction based on business semantics. It can only mechanically load all models with monitoring tags and cannot automatically optimize the assembly effect based on complex factors such as the tower's mounting height and obstruction conditions.

[0030] (4) Redundancy in visual interaction When loading large amounts of resources, UI panels severely overlap and obscure each other in three-dimensional space, even obscuring core entities. The lack of a dynamic adaptive layout mechanism seriously affects the user's information acquisition efficiency and decision-making experience.

[0031] To address the aforementioned technical issues, this application proposes a method and apparatus for constructing a 3D twin scene, establishing an automated generation framework based on semantic-driven, declarative rules, and dynamic assembly to solve problems such as non-automated processes, hard-coded logic, business semantic gaps, and redundant visual interactions in existing technologies.

[0032] Figure 1 This is a flowchart illustrating a method for constructing a three-dimensional twin scene provided in an embodiment of this application, as shown below. Figure 1 As shown, the method includes the following steps: S1. Construct a semantic ontology model to map physical assets into semantic nodes that include business capabilities, real-time status, and logical relationships.

[0033] Furthermore, physical tower assets are defined as semantic nodes that include business capabilities, real-time status, and logical relationships. Business capabilities can include monitoring capabilities and power supply levels. Logical relationships can include neighboring topologies.

[0034] S2. Based on the semantic ontology model, in response to the user's spatial operations on the 3D digital twin platform, the corresponding business intent tag is captured only when the physical asset pointed to by the spatial operation has been mapped as a semantic node, and the business intent tag is converted into a structured query instruction.

[0035] Optionally, the user's spatial operations on the 3D digital twin base can be polygon selection; capturing the corresponding business intent tags can generate tags for fire prevention scenarios.

[0036] This application can convert the aforementioned fire prevention scenario generated tags into structured query instructions.

[0037] S3. Based on the structured query instructions, automatically index and assemble the corresponding 3D models, dynamic visual components and interactive panels from the predefined visualization asset library; calculate the non-occluded arrangement position of the interactive panels in 3D space in real time based on the camera view and the entity bounding box.

[0038] It should be noted that this application uses declarative rules to automatically assemble operators, employing condition-action logic. When a specific business intent is met, the corresponding 3D model (tower type), dynamic visual components (coverage area), and interactive UI (data panel) are automatically indexed from the visual asset library. Based on camera dynamic parameters (viewpoint, zoom) and entity bounding boxes, the non-occluded arrangement position of the UI panel in 3D space is calculated in real time.

[0039] This application proposes a method for constructing 3D twin scenes. By establishing an automated generation framework that combines semantic-driven, declarative rules, and dynamic assembly, it can solve problems such as non-automated processes, hard-coded logic, business semantic gaps, and redundant visual interactions in existing technologies.

[0040] For example, in one instance, in response to a user performing a selection action on a 3D digital twin pedestal, the system records a set of coordinates. G (Geofencing), and receiving business intent tags. I (For example: I =Emergency Support). System Tag Parsing I The system retrieves a set of entities E with the corresponding capabilities from the semantic ontology model. For example, it filters out all iron towers in region G that have a backup power duration of >4 hours and are equipped with emergency microwaves.

[0041] The system's assembly operator performs operations according to declarative rules: Action 1: Load the high-precision 3D model of the selected entity E; Action 2: Automatically attach special effects (such as flashing red pulse effect in alarm state) according to the entity's status; Action 3: Dynamically call the preset business dashboard component and bind it to the entity's real-time monitoring data stream via WebSocket.

[0042] Based on the current viewpoint V obtained by the system processor, to prevent the data panels of multiple towers from overlapping in screen space, the system calculates the coordinates of the panel's center point: Where P represents the three-dimensional coordinates of the center point of the data panel of the tower, O represents the three-dimensional coordinates of the center position of the corresponding tower body, and n is the normal vector of the line of sight pointing to the tower from the current camera viewpoint, that is, the direction in which the user's eyes look at the tower (from the screen to the tower). The vector representing the vertically upward direction is s, which is the offset step size dynamically calculated based on the panel crowding level within the scene. This method ensures that the panels neither obstruct the tower nor overlap each other.

[0043] In this application, non-developers (such as business managers and front-line operations personnel) do not need to have 3D visualization development experience; they can create complex digital twin scenes simply by defining business rules. The system automatically maps business rules into assembly instructions for 3D components, realizing a direct transformation from business requirements to 3D scenes, completely changing the lengthy chain of traditional business requirements → developer coding implementation.

[0044] Furthermore, in this application, the scene content is no longer a preset static template, but is dynamically generated based on current business metrics (such as real-time status data like abnormal temperature readings, power outage alarms, and backup power duration). The system can perceive changes in business data in real time and automatically trigger scene updates and reconstructions, ensuring that users always see accurate information highly relevant to the current business situation, rather than redundant full-scale data.

[0045] This application also completely solves the visual clutter problem commonly found in automatically generated scenarios by introducing an intelligent obstacle avoidance layout algorithm based on camera perspective and entity bounding boxes. In dense scenarios with large-scale resource loading, the UI panel can automatically find the optimal floating position, neither obscuring core entities nor overlapping each other, ensuring that all business information is clearly visible, greatly improving the user's information acquisition efficiency and decision-making experience.

[0046] This application abandons the fixed model of pre-developed complete templates in traditional solutions, and adopts an assembly method of atomic-level components and real-time combination. Components such as 3D models, visual effects, and data panels can be freely combined and dynamically superimposed according to business intentions, no longer limited by the number and form of preset templates. Whether it is a single business scenario (such as emergency support) or a complex business scenario (such as emergency support + forest fire prevention + drone dispatch), this application can instantly assemble a suitable sandbox through a rule engine, truly realizing the infinite expansion of business intentions.

[0047] In this embodiment of the application, a semantic ontology model is constructed to map physical assets into semantic nodes that include business capabilities, real-time status, and logical relationships, including the following steps: S4. Abstract physical assets into ontology classes, and define static slots for each ontology class to describe its inherent attributes and dynamic slots to describe its functional characteristics.

[0048] S5. Extract the attribute data of specific physical assets from the underlying resource database, fill them into the static and dynamic slots of the corresponding ontology class, and generate semantic node instances that correspond one-to-one with the physical assets.

[0049] S6. Based on the spatial topology and business logic relationships between physical assets, construct object attribute associations between semantic node instances to form a network structure with business semantic associations.

[0050] In this embodiment, in response to a user's spatial operation on a 3D digital twin platform, the corresponding business intent tag is captured only when the physical asset pointed to by the spatial operation has been mapped as a semantic node, and the business intent tag is converted into a structured query instruction, including: S7. In response to the user's spatial operation on the 3D digital twin base, determine whether the physical asset pointed to by the spatial operation has been mapped as a semantic node; if not, terminate the process; if so, extract the geofence information corresponding to the spatial operation and encapsulate the geofence information into a spatial query predicate.

[0051] S8. If the mapping exists, obtain the business intent tag specified by the user, perform semantic parsing on the business intent tag according to the preset business semantic dictionary, and map the business intent tag to at least one business capability predicate.

[0052] S9. Logically combine the spatial query predicate with the business capability predicate to generate a structured query instruction that conforms to the preset syntax.

[0053] In this embodiment of the application, based on structured query instructions, the corresponding 3D model, dynamic visual components, and interactive panel are automatically indexed and assembled from a predefined visual asset library, including: S11. Receive and parse the structured query command, extract the business intent tag and entity type information in the command, and index the 3D model template and visual component template that match the business intent from the visualization asset library according to the preset business intent-visual component mapping table.

[0054] S12. Obtain the real-time status data of each entity in the entity set corresponding to the structured query command, match the real-time status data with the condition part of the preset condition-action rule set, and filter out the entities that need to trigger special visual performance.

[0055] S13. Based on the index results and matching results, perform the assembly operation, which includes: instantiating the 3D model template to generate a 3D model instance and attaching it to the corresponding entity position; instantiating the visual component template to generate a dynamic visual component and associating it with the corresponding entity; and dynamically calling and binding the corresponding interactive panel for each entity.

[0056] In this embodiment of the application, the non-occluded arrangement position of the interactive panel in three-dimensional space is calculated in real time based on the camera view and the physical bounding box, including: S14. Obtain the view matrix and projection matrix of the current camera view, project the 3D bounding box of each entity and the candidate position of each interactive panel onto the screen coordinate system, and generate the occupied area of ​​each entity in the screen space and the candidate area of ​​each interactive panel in the screen space.

[0057] S15. Calculate the degree of occlusion between the candidate area of ​​each interactive panel and the area occupied by the corresponding entity, as well as the degree of overlap between the candidate areas of each interactive panel. When the degree of occlusion or overlap exceeds the preset threshold, trigger the position adjustment mechanism.

[0058] S16. Using a physical simulation-based repulsion model or a gradient descent-based optimization algorithm, with the goal of minimizing the degree of occlusion and overlap, iteratively adjust the position offset vector of each interactive panel in three-dimensional space until the preset visual clarity constraints are met.

[0059] In this embodiment, the degree of occlusion between the candidate area of ​​each interactive panel and the area occupied by the corresponding entity, as well as the degree of overlap between the candidate areas of each interactive panel, are calculated. When the degree of occlusion or overlap exceeds a preset threshold, a position adjustment mechanism is triggered, including: S17. Project the 3D bounding box of each entity onto the screen space to generate the polygon of the entity's occupied area. Project the candidate position of each interactive panel onto the screen space to generate the polygon of the panel candidate area. Calculate the intersection-union ratio between the panel candidate area and the corresponding entity's occupied area, and determine the intersection-union ratio as the occlusion degree value.

[0060] S18. Obtain the candidate region polygons of all interactive panels within the same screen space, traverse and calculate the intersection-union ratio between every two different interactive panel candidate regions, and determine the maximum value or cumulative value of all intersection-union ratios as the degree of overlap.

[0061] S19. Compare the occlusion level value with the preset occlusion threshold and the overlap level value with the preset overlap threshold. If the occlusion level value exceeds the occlusion threshold or the overlap level value exceeds the overlap threshold, it is determined that there is a visual conflict and the position adjustment mechanism is triggered.

[0062] In this embodiment, a repulsive force model based on physical simulation or an optimization algorithm based on gradient descent is used to iteratively adjust the position offset vector of each interactive panel in three-dimensional space, with the goal of minimizing the degree of occlusion and overlap, until a preset visual clarity constraint is met, including: S21. Construct a total energy function with the position offset vectors of all interactive panels as independent variables. The total energy function includes an occlusion penalty term, an overlap penalty term, and a position constraint term. The occlusion penalty term is used to quantify the degree of occlusion between the interactive panel and the corresponding entity. The overlap penalty term is used to quantify the degree of overlap between the interactive panels. The position constraint term is used to limit the distance of the interactive panel from its initial anchor position.

[0063] The weight coefficients of the occlusion penalty term, overlap penalty term, and position constraint term in the total energy function can be adaptively adjusted according to the needs of the actual business scenario. This embodiment does not limit their fixed values.

[0064] S22. Using an optimization algorithm based on gradient descent or a repulsive force model based on physical simulation, with the goal of minimizing the total energy function, iteratively update the position offset vector of each interactive panel. In each iteration, adjust the offset amount according to the gradient direction of the offset vector based on the current total energy function, or adjust the offset amount according to the sum of the repulsive force between panels and the attraction force of the entity on the panel.

[0065] This application can suppress the oscillation of position offset by setting a damping coefficient during the iteration process, thereby further ensuring the convergence stability of the algorithm.

[0066] S23. After each iteration update, recalculate the current total energy function value. When the total energy function value is less than the preset energy threshold, or the change in the energy function value of multiple consecutive iterations is less than the preset change threshold, or the number of iterations reaches the preset maximum number of iterations, stop the iteration and output the current position offset vector as the final arrangement result.

[0067] Figure 2 A block diagram of a three-dimensional twin scene construction device provided in the embodiments of this application, such as... Figure 2 As shown, the system includes: an ontology construction module, a structure query module, and an automatic assembly module. The ontology construction module builds a semantic ontology model, mapping physical assets to semantic nodes that contain business capabilities, real-time status, and logical relationships. The structure query module, based on the semantic ontology model, responds to user spatial operations on the 3D digital twin platform. Only when the physical asset pointed to by the spatial operation has been mapped to a semantic node, it captures the corresponding business intent tag and converts it into a structured query instruction. The automatic assembly module automatically indexes and assembles corresponding 3D models, dynamic visual components, and interactive panels from a predefined visualization asset library based on the structured query instructions; it calculates the unobstructed arrangement position of the interactive panels in 3D space in real time based on the camera viewpoint and entity bounding boxes.

[0068] Based on the above embodiments, the automatic assembly module includes: a declarative rule-based automatic assembly operator and a visual avoidance layout processor. The declarative rule-based automatic assembly operator employs condition-action logic. When a specific business intent is met, it automatically indexes the corresponding 3D model (tower type), dynamic visual components (coverage area), and interactive UI (data panel) from the visual asset library.

[0069] Figure 3 This is a logical architecture diagram of a three-dimensional twin scene construction device provided in an embodiment of this application. Figure 3 As shown in the example, the 3D twin scene construction system includes four core hierarchical structures: data foundation layer, business semantic modeling layer, intent parsing and rule engine layer, and dynamic assembly and rendering layer.

[0070] The data foundation layer includes physical assets such as towers and data centers, along with their attribute databases. The business semantic modeling layer establishes entity ontology models and business rule graphs, enabling machines to understand business concepts such as fire prevention and emergency response.

[0071] The intent parsing and rule engine layer includes a visual interactive interface that receives the user's selected area and intent label; the intent parser converts the intent into instructions; and the declarative rule engine generates rules based on logic.

[0072] Dynamic Assembly and Rendering Layer: The dynamic assembly operator calls the visual asset library, the layout processor applies an anti-occlusion algorithm, and finally generates a customized 3D business scene.

[0073] Figure 4 This is a schematic diagram illustrating the automatic generation of a three-dimensional twin scene, as provided in an embodiment of this application. Figure 4 As shown, it includes the following steps: S10: Intent Capture. Responds to the user selecting an irregular area in the 3D interface and choosing the intent label for forest fire prevention.

[0074] S20: Semantic matching. The system sends the region and intent to the backend ontology model for semantic querying, and matches entities with firewall capabilities.

[0075] S30: Rule parsing. The rule engine executes instructions to determine the model components to be loaded, the coverage radius, and the data panel binding relationships.

[0076] S40: Dynamic assembly and data binding. The scene assembly operator calls components from the asset library, the layout processor calculates the unobstructed positions of the panel, and finally generates a customized 3D business sandbox.

[0077] Figure 5 A comparison diagram showing the performance of an intelligent obstacle avoidance algorithm provided in an embodiment of this application. For example... Figure 5 As shown, Figure 5 The image on the left shows the existing technology, which uses a traditional coordinate anchoring method. When a large number of towers are loaded, the data panels above them become misaligned, resulting in severe overlap, misalignment, and obstruction, making it impossible to clearly display information.

[0078] Figure 5 The figure on the right shows the solution of this application. This application introduces an intelligent layout processor, which can calculate a visual anti-occlusion spatial network based on the camera view parameters. In addition, this application uses an adaptive anchoring algorithm to enable all data panels to automatically find the optimal non-overlapping floating position in 3D space (as shown by the numerical offsets 1 and 2 in the figure), so as to ensure that all information is clearly visible and the anchoring relationship with the object is accurate.

[0079] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0080] Figure 6 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application, such as... Figure 6 As shown, the electronic device includes: a processor 310, a communication interface 320, a memory 330 for storing processor-executable computer programs, and a communication bus 340. The processor 310, communication interface 320, and memory 330 communicate with each other via the communication bus 340. The processor 310 executes the executable computer program to implement the steps of the aforementioned three-dimensional twin scene construction method.

[0081] It is understandable that, in addition to memory and a processor, this electronic device may also include input devices such as a keyboard, output devices such as a display, and other communication modules. The input devices, output devices, and other communication modules communicate with the processor through I / O interfaces (i.e., input / output interfaces).

[0082] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof. These programming languages ​​include, but are not limited to, object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0083] Furthermore, this application also discloses a computer-readable storage medium, which, when the instructions in the computer-readable storage medium are executed by the processor of an electronic device, enables the electronic device to perform the various steps of the three-dimensional twin scene construction method disclosed in this application.

[0084] In the context of this application, computer-readable storage media can be tangible media, and more specific examples include portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, and portable compact disc read-only memory (CD). ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0085] Specifically, according to embodiments of this application, the processes described in the flowcharts can be implemented as computer software programs. For example, embodiments of this application include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the three-dimensional twin scene construction method disclosed in this application. When the computer program is executed by a processing device, it performs the functions defined in the methods of the embodiments of this application.

[0086] While the foregoing discussion includes several specific implementation details, these should not be construed as limiting the scope of this application. The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of disclosure in this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described concept.

[0087] Those skilled in the art should also understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for constructing a three-dimensional twin scene, characterized in that, The method includes: Construct a semantic ontology model to map physical assets into semantic nodes that contain business capabilities, real-time status, and logical relationships; Based on the semantic ontology model, in response to the user's spatial operations on the three-dimensional digital twin base, the corresponding business intent tag is captured only when the physical asset pointed to by the spatial operation has been mapped as a semantic node, and the business intent tag is converted into a structured query instruction. Based on the structured query instructions, the corresponding 3D models, dynamic visual components, and interactive panels are automatically indexed and assembled from the predefined visualization asset library; based on the camera view and the entity bounding box, the unobstructed arrangement position of the interactive panels in 3D space is calculated in real time.

2. The method according to claim 1, characterized in that, Construct a semantic ontology model that maps physical assets to semantic nodes containing business capabilities, real-time status, and logical relationships, including: Abstract physical assets into ontology classes, and define static slots for describing their inherent attributes and dynamic slots for describing their functional characteristics for each ontology class. Extract the attribute data of specific physical assets from the underlying resource database, populate the static and dynamic slots of the corresponding ontology class, and generate semantic node instances that correspond one-to-one with the physical assets. Based on the spatial topological and business logic relationships between physical assets, object attribute associations are constructed between semantic node instances to form a network structure with business semantic associations.

3. The method according to claim 1, characterized in that, In response to a user's spatial operations on a 3D digital twin platform, only when the physical asset targeted by the spatial operation has been mapped to a semantic node, the corresponding business intent tag is captured, and the business intent tag is converted into a structured query instruction, including: In response to a user's spatial operation on the 3D digital twin base, it is determined whether the physical asset to which the spatial operation is directed has been mapped as a semantic node; if not, the process is terminated; if so, the geofence information corresponding to the spatial operation is extracted and the geofence information is encapsulated into a spatial query predicate. If the mapping exists, obtain the business intent tag specified by the user, and perform semantic parsing on the business intent tag according to the preset business semantic dictionary, and map the business intent tag to at least one business capability predicate; The spatial query predicate and the business capability predicate are logically combined to generate a structured query instruction that conforms to a preset syntax.

4. The method according to claim 1, characterized in that, Based on the structured query instructions, the system automatically indexes and assembles corresponding 3D models, dynamic visual components, and interactive panels from a predefined visualization asset library, including: Receive and parse the structured query instruction, extract the business intent tag and entity type information in the instruction, and index the 3D model template and visual component template that match the business intent from the visualization asset library according to the preset business intent-visual component mapping table; Obtain the real-time status data of each entity in the entity set corresponding to the structured query instruction, match the real-time status data with the condition part of the preset condition-action rule set, and filter out the entities that need to trigger special visual performance. Based on the index results and matching results, an assembly operation is performed, which includes: instantiating the 3D model template to generate a 3D model instance and attaching it to the corresponding entity position; instantiating the visual component template to generate a dynamic visual component and associating it with the corresponding entity; and dynamically calling and binding the corresponding interactive panel for each entity.

5. The method according to claim 1, characterized in that, Based on the camera viewpoint and the physical bounding box, the unobstructed arrangement position of the interactive panel in three-dimensional space is calculated in real time, including: Obtain the view matrix and projection matrix of the current camera view, project the 3D bounding box of each entity and the candidate position of each interactive panel onto the screen coordinate system, and generate the occupied area of ​​each entity in the screen space and the candidate area of ​​each interactive panel in the screen space. Calculate the degree of occlusion between the candidate area of ​​each interactive panel and the area occupied by the corresponding entity, as well as the degree of overlap between the candidate areas of each interactive panel. When the degree of occlusion or overlap exceeds a preset threshold, trigger the position adjustment mechanism. Based on a physical simulation-based repulsion model or a gradient descent-based optimization algorithm, the position offset vectors of each interactive panel in three-dimensional space are iteratively adjusted to minimize the degree of occlusion and overlap until the preset visual clarity constraints are met.

6. The method according to claim 5, characterized in that, Calculate the degree of occlusion between the candidate areas of each interactive panel and the corresponding entity-occupied areas, as well as the degree of overlap between the candidate areas of each interactive panel. When the degree of occlusion or overlap exceeds a preset threshold, trigger a position adjustment mechanism, including: Project the 3D bounding box of each entity onto the screen space to generate the polygon of the area occupied by the entity. Project the candidate position of each interactive panel onto the screen space to generate the polygon of the candidate area of ​​the panel. Calculate the intersection-union ratio between the candidate area of ​​the panel and the area occupied by the corresponding entity, and determine the intersection-union ratio as the occlusion degree value. Obtain candidate region polygons for all interactive panels within the same screen space, iterate through and calculate the intersection-union ratio between every two candidate regions of different interactive panels, and determine the maximum value or cumulative value of all intersection-union ratios as the degree of overlap. The occlusion degree value is compared with a preset occlusion threshold, and the overlap degree value is compared with a preset overlap threshold. If the occlusion degree value exceeds the occlusion threshold or the overlap degree value exceeds the overlap threshold, it is determined that there is a visual conflict, and the position adjustment mechanism is triggered.

7. The method according to claim 5, characterized in that, Based on a physical simulation-based repulsion model or a gradient descent-based optimization algorithm, the position offset vectors of each interactive panel in three-dimensional space are iteratively adjusted to minimize the degree of occlusion and overlap until preset visual clarity constraints are met, including: Construct a total energy function with the position offset vectors of all interactive panels as independent variables. The total energy function includes an occlusion penalty term, an overlap penalty term, and a position constraint term. The occlusion penalty term is used to quantify the degree of occlusion between the interactive panel and the corresponding entity. The overlap penalty term is used to quantify the degree of overlap between the interactive panels. The position constraint term is used to limit the distance of the interactive panel from its initial anchor position. The position offset vectors of each interactive panel are iteratively updated with the goal of minimizing the total energy function by using an optimization algorithm based on gradient descent or a repulsive force model based on physical simulation. In each iteration, the offset is adjusted according to the gradient direction of the current total energy function on the offset vector, or according to the sum of the repulsive force between panels and the attraction force of the entity on the panel. After each iteration, the current total energy function value is recalculated. When the total energy function value is less than the preset energy threshold, or the change in the energy function value after multiple consecutive iterations is less than the preset change threshold, or the number of iterations reaches the preset maximum number of iterations, the iteration stops and the current position offset vector is output as the final arrangement result.

8. A three-dimensional twin scene construction device, characterized in that, include: The ontology module is used to build a semantic ontology model, mapping physical assets into semantic nodes that contain business capabilities, real-time status, and logical relationships. The structure query module is used to respond to the user's spatial operations on the three-dimensional digital twin base based on the semantic ontology model. Only when the physical asset pointed to by the spatial operation has been mapped as a semantic node, the module captures the corresponding business intent tag and converts the business intent tag into a structured query instruction. The automatic assembly module is used to automatically index and assemble the corresponding 3D model, dynamic visual components and interactive panels from the predefined visual asset library according to the structured query instructions; and to calculate the non-occluded arrangement position of the interactive panels in 3D space in real time based on the camera view and the entity bounding box.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the method as described in any one of claims 1 to 7.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method as described in any one of claims 1 to 7.

Images