Method, device and equipment for intelligently controlling BIM model and storage medium

Abstract

The invention discloses a method, device and equipment for intelligently controlling a BIM model and a storage medium, and the method comprises the steps: receiving a natural language interaction instruction for a target BIM model, and converting the natural language interaction instruction into a structured instruction comprising a target building object and an interaction intention through a preset building domain knowledge base; if the interaction intention is a query intention, obtaining related model data about the target building object from the target BIM model through a data query tool, and forming query result information meeting the query intention based on the related model data; if the interaction intention is a control intention, firstly determining identity recognition information of the target building object from the target BIM model through a data query tool, and then generating and executing a control code corresponding to the control intention and the identity recognition information through a model control tool so as to implement specified operation on the target building object; according to the invention, the interaction process between the user and the BIM model is simplified.

Description

Technical Field

[0001] This invention relates to the field of computer-aided design technology, and in particular to a method, apparatus, device, and storage medium for intelligent control of BIM models. Background Technology

[0002] With the widespread adoption of BIM (Building Information Modeling) in the engineering construction field, BIM models have become the core carrier for the full lifecycle management of building projects. However, the current 3D model review process suffers from significant efficiency bottlenecks, particularly the contradiction between the complexity of 3D view navigation and the difficulty for review experts to quickly locate problematic components. In the process of reviewing 3D drawings using existing BIM platforms, users rely on complex menu operations and parameter settings to complete model queries, significantly increasing the barrier to entry for non-professionals. Review experts spend a considerable amount of time on view switching and component searching, rather than substantive compliance checks. This operational inefficiency directly restricts the efficiency of the 3D model review process. Therefore, how to reduce the complexity of user operation of BIM models to improve the efficiency of the 3D model review process has become a pressing technical problem that needs to be solved by those skilled in the art. Summary of the Invention

[0003] The purpose of this invention is to provide a method, apparatus, device, and storage medium for intelligent control of BIM models, which simplifies the interaction process between users and BIM models to lower the technical threshold and achieve a dual improvement in the efficiency and accuracy of model review.

[0004] According to one aspect of the present invention, a method for intelligent control of a BIM model is provided, the method comprising: Receive natural language interaction instructions for the target BIM model, and use a preset building domain knowledge base to convert the natural language interaction instructions into structured instructions that include the target building object and the interaction intent; If the interaction intent is a query intent, then relevant model data about the target building object is obtained from the target BIM model through the data query tool, and query result information that satisfies the query intent is formed based on the relevant model data; If the interaction intent is a control intent, the identification information of the target building object is first determined from the target BIM model by the data query tool, and then the control code corresponding to the control intent and the identification information is generated and executed by the model control tool to perform the specified operation on the target building object.

[0005] Optionally, the step of converting the natural language interaction instructions into structured instructions including the target building object and the interaction intent using a preset architectural domain knowledge base includes: The natural language interaction commands are parsed to extract building object nouns, building object attribute nouns, component nouns, and interaction verbs; Determine whether the architectural object name exists in the business object document in the architectural domain knowledge base. If so, obtain the object explanation information associated with the architectural object name from the business object document, and form the target architectural object based on the architectural object name and the object explanation information. Using the object attribute documents in the aforementioned architectural knowledge base, the architectural object attribute names are converted into standard attribute items; Using the basic component documents in the aforementioned architectural knowledge base, the component names are converted into standard component types; Using the interaction intent documents in the aforementioned architectural knowledge base, the interaction verbs are converted into standard interaction types; The interaction intent is formed based on the standard attribute items, the standard component types, and the standard interaction types.

[0006] Optionally, if the interaction intent is a query intent, then relevant model data about the target building object is obtained from the target BIM model through a data query tool, and query result information that satisfies the query intent is formed based on the relevant model data, including: When the standard interaction type in the query intent is component and attribute query, a first component belonging to the target building object and satisfying the standard component type is determined from the target BIM model, and the attribute values ​​of the first component with respect to the standard attribute items are obtained. The query result information is then formed based on the first component and the attribute values; or, When the standard interaction type in the query intent is a collision query, a second component belonging to the target building object and satisfying the standard component type is determined from the target BIM model, and collision components that have a collision relationship with the second component are determined. The query result information is then formed based on the second component and the collision components; or, When the standard interaction type in the query intent is a connectivity query, a third component belonging to the target building object and satisfying the standard component type is determined from the target BIM model, and a connecting component that has a connectivity relationship with the third component is determined. The query result information is then formed based on the third component and the connecting component.

[0007] Optionally, the step of generating and executing control code corresponding to the control intent and the identity information through the model manipulation tool to perform a specified operation on the target building object includes: The model manipulation tool obtains a target operation interface that matches the manipulation intention from a preset operation interface specification; The model manipulation tool combines the identity information with the target operation interface to generate control code that conforms to a predefined syntax specification. The control code is executed by the model manipulation tool to perform specified operations on the target building object.

[0008] Optionally, the method further includes: Based on the query results and / or the operation results of performing the specified operation, generate visual interactive results. The interaction result information includes at least one of the following: text description data, structured data table, and model view corresponding to the control code.

[0009] To achieve the above objectives, the present invention also provides an apparatus for intelligently controlling a BIM model, the apparatus comprising: The conversion module is used to receive natural language interaction instructions for the target BIM model and convert the natural language interaction instructions into structured instructions including the target building object and the interaction intent using a preset building domain knowledge base. The query module is used to obtain relevant model data about the target building object from the target BIM model through a data query tool if the interaction intent is a query intent, and to form query result information that satisfies the query intent based on the relevant model data; The control module is used to, if the interaction intent is a control intent, first determine the identity information of the target building object from the target BIM model through the data query tool, and then generate and execute control code corresponding to the control intent and the identity information through the model control tool to perform a specified operation on the target building object.

[0010] Optionally, the control module is used for: The model manipulation tool obtains a target operation interface that matches the manipulation intention from a preset operation interface specification; The model manipulation tool combines the identity information with the target operation interface to generate control code that conforms to a predefined syntax specification. The control code is executed by the model manipulation tool to perform specified operations on the target building object.

[0011] Optionally, the device further includes a display module for: Based on the query results and / or the operation results of performing the specified operation, generate visual interactive results. The interaction result information includes at least one of the following: text description data, structured data table, and model view corresponding to the control code.

[0012] To achieve the above objectives, the present invention also provides a computer device, which specifically includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the intelligent control BIM model method described above.

[0013] To achieve the above objectives, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method for intelligent control of BIM models described above.

[0014] The present invention provides a method, apparatus, device, and storage medium for intelligent control of BIM models. By receiving natural language interaction commands and automatically converting them into structured commands, it achieves intelligent parsing and control of BIM models. When a query intent is recognized, the system automatically retrieves and returns the required model data from the BIM model; when a manipulation intent is recognized, it automatically generates and executes control code to complete the operation on the target object through a query-then-manipulation process. This invention enables users to accurately control the BIM platform using everyday language without requiring specialized BIM platform operation skills. It lowers the barrier to entry, ensures accurate identification of user needs through structured command conversion, and guarantees the accuracy and efficiency of operation through targeted query and manipulation processing logic, significantly improving the convenience and practicality of BIM model interaction. Attached Figure Description

[0015] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 This is an optional flowchart illustrating the method for intelligent control of a BIM model provided in Embodiment 1. Figure 2 This is a schematic diagram of an optional component structure of the device for intelligent control of BIM models provided in Embodiment 2. Figure 3 This is a schematic diagram of an optional hardware structure for the computer device provided in Embodiment 4. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without inventive effort are within the scope of protection of this invention.

[0017] Example 1 This invention provides a method for intelligent control of BIM models, such as... Figure 1 As shown, the method specifically includes the following steps: Step S101: Receive natural language interaction instructions for the target BIM model, and use a preset building domain knowledge base to convert the natural language interaction instructions into structured instructions that include the target building object and the interaction intent.

[0018] The natural language interaction commands are commands input by users (such as review experts) in everyday language to control the target BIM model. These commands may include voice interaction commands and / or text interaction commands, such as "show the bathroom on the second floor" or "highlight all concrete columns".

[0019] The aforementioned architectural knowledge base is a pre-built collection that stores architectural terms, aliases, explanations, and mapping relationships between them. The architectural knowledge base can provide a basis for the conversion between natural language and BIM model recognizable standard information. Through the mapping relationships in the architectural knowledge base, everyday language in natural language interaction instructions can be converted into standardized professional terms.

[0020] The target building object can be a building space object pointed to in the user interaction command, such as a bathroom, stairwell, electrical room, etc.; or it can be a building component or equipment object, such as a fire hydrant, door, pipe, etc.

[0021] The interaction intent is a standardized description of the user's core interaction needs, including at least the interaction intent type (such as query intent or manipulation intent) and interaction requirement parameters (such as the attribute query requirement corresponding to the query intent, the operation type requirement corresponding to the manipulation intent, etc.).

[0022] The structured instructions convert non-standardized information in natural language interaction instructions into standardized instructions that can be recognized by the BIM platform, laying the foundation for subsequent operations.

[0023] Step S102: If the interaction intent is a query intent, then relevant model data about the target building object is obtained from the target BIM model through the data query tool, and query result information that satisfies the query intent is formed based on the relevant model data.

[0024] It should be noted that the query intent is only to obtain relevant model information rather than to change the model state. For example, the query intent may be to query the attribute information of the target building object, the quantity of the target building object, or the relationship between the target building object and other objects.

[0025] The data query tool is a pre-configured functional module used to extract or calculate required information from BIM model data sources. It supports standardized data interaction with BIM models and can perform specialized query logic such as attribute filtering, collision detection, and connectivity analysis. Furthermore, the relevant model data refers to BIM model data directly associated with the target building object, reflecting its own characteristics or relationships. For example, when the target building object is a "second-floor bathroom," the relevant model data may include the bathroom's area, the floor it belongs to, and the associated door / window component identifiers.

[0026] Step S103: If the interaction intent is a control intent, the identity information of the target building object is first determined from the target BIM model by the data query tool, and then the control code corresponding to the control intent and the identity information is generated and executed by the model control tool to perform the specified operation on the target building object.

[0027] The control intent refers to changing the display or functional state of the target building object in the target BIM model. For example, the control intent could be: highlighting a component (e.g., highlighting a park gate), modifying its color (e.g., marking a fire hydrant in red), or focusing the view (e.g., focusing the view on a stairwell). Furthermore, the model control tool is a code generator built based on the application programming interface (API) specification of a BIM model management platform (e.g., BIMFace). It can dynamically generate executable control code according to operational requirements. This control code is an executable script conforming to the syntax specification of the BIM model management platform. By executing the control code, specified APIs in the BIM model management platform can be called to perform specified operations on the target building object.

[0028] In this embodiment, the model manipulation tool can automatically generate control code that can be used by the BIM model management platform. The model manipulation tool executes the control code by calling the BIM model management platform and collects the code execution result information.

[0029] In this embodiment, intelligent parsing and control of the BIM model is achieved by receiving natural language interaction commands and automatically converting them into structured commands. When a query intent is recognized, the system can automatically retrieve and return the required model data from the BIM model; when a manipulation intent is recognized, control code is automatically generated and executed to complete the operation on the target object through a query-then-manipulation process. This achieves the effect of allowing users to accurately control the BIM platform using everyday language without needing professional BIM platform operation skills. This lowers the barrier to entry, ensures accurate identification of user needs through structured command conversion, and guarantees the accuracy and efficiency of operation through targeted query and manipulation processing logic, significantly improving the convenience and practicality of BIM model interaction.

[0030] Specifically, step S101, which involves using a preset architectural domain knowledge base to convert the natural language interaction command into a structured command that includes the target architectural object and the interaction intent, includes the following steps: Step A1: Parse the building object nouns, building object attribute nouns, component nouns, and interaction verbs from the natural language interaction instructions; It should be noted that a building object refers to a building space or building facility, while a component is a basic unit contained in or constituting a building object.

[0031] Step A2: Determine whether the architectural object term exists in the business object document in the architectural domain knowledge base. If so, obtain the object explanation information associated with the architectural object term from the business object document, and form the target architectural object based on the architectural object term and the object explanation information. The business object document includes: object type (such as building space or building facility), building object name, and object explanation information. The object explanation information is used to expand the building object name to facilitate subsequent data query operations. For example, if the object type is "building space" and the building object name is "1st floor", then the object explanation information is "belongs to one of the following floors: 1F, F1, 1st floor, first floor, 1, 1st floor"; if the object type is "building space" and the building object name is "toilet", then the object explanation information is "the room name includes one of the following: bathroom, WC, toilet, washbasin, handwashing"; in addition, if the building object name does not exist in the business object document, then the building object name is directly set as the target building object.

[0032] Step A3: Using the object attribute documents in the architecture domain knowledge base, convert the architectural object attribute names into standard attribute items; The object attribute document includes: standard attribute items (such as element number) and attribute aliases (such as ID, number); if the attribute name of the building object belongs to the attribute alias in the object attribute document, it is mapped to the corresponding standard attribute item.

[0033] Step A4: Using the basic component documents in the building domain knowledge base, convert the component names into standard component types; The basic component document includes: standard component types (such as structural walls) and component aliases (such as walls, wall bodies); if the component name belongs to a component alias in the basic component document, it is mapped to the corresponding standard component type.

[0034] Step A5: Using the interaction intent document in the architecture domain knowledge base, convert the interaction verbs into standard interaction types; The interaction intent document includes standard interaction types (such as collision query) and everyday verbs (such as pass through, traverse); if the interaction verb belongs to the everyday verbs in the interaction intent document, it is mapped to the corresponding standard interaction type.

[0035] Step A6: Form the interaction intent based on the standard attribute items, the standard component type, and the standard interaction type.

[0036] It should also be noted that during the instruction mapping and conversion process using the architectural domain knowledge base, if the architectural object nouns, architectural object attribute nouns, component nouns, and interaction verbs parsed in step A1 do not exist in the architectural domain knowledge base, then there is no need for excessive interpretation. The architectural object nouns are directly set as the target architectural object, and the architectural object attribute nouns, component nouns, and interaction verbs are set as interaction intents.

[0037] In practical applications, at least architectural object nouns and interaction verbs are parsed from the natural language interaction instructions. The architectural domain knowledge base is then used to convert the architectural object nouns into target architectural objects and the interaction verbs into interaction intents. Additionally, when architectural object attribute names and / or component names are also parsed from the natural language interaction instructions, the architectural domain knowledge base is used to convert these parsed architectural object attribute names and / or component names into interaction intents as well. Therefore, interaction intents at least include standard interaction types converted from interaction verbs, and may also include standard attribute items converted from architectural object attribute names and / or standard component types converted from component names.

[0038] In this embodiment, the collaborative work of multiple professional documents in the architectural knowledge base enables accurate conversion from user natural language to standardized system instructions. Specifically, the "business object document" not only converts user-input architectural object terms into standard terminology, but more importantly, it semantically expands them to generate attribute filtering conditions containing multiple relevant keywords. This expansion mechanism effectively broadens the query scope and avoids query omissions caused by non-standard model data annotation or differences in user terminology.

[0039] Furthermore, step S102 specifically includes the following steps: When the standard interaction type in the query intent is component and attribute query, a first component belonging to the target building object and satisfying the standard component type is determined from the target BIM model, and the attribute values ​​of the first component with respect to the standard attribute items are obtained. The query result information is then formed based on the first component and the attribute values; or, When the standard interaction type in the query intent is a collision query, a second component belonging to the target building object and satisfying the standard component type is determined from the target BIM model, and collision components that have a collision relationship with the second component are determined. The query result information is then formed based on the second component and the collision components; or, When the standard interaction type in the query intent is a connectivity query, a third component belonging to the target building object and satisfying the standard component type is determined from the target BIM model, and a connecting component that has a connectivity relationship with the third component is determined. The query result information is then formed based on the third component and the connecting component.

[0040] In this embodiment, when the interaction intent is a query, three types of query methods are supported: component and attribute query, to retrieve the required component list and component attribute values; collision query, to retrieve collision components that can directly collide with the target component in 3D space or collide after transformation; and connectivity query, to retrieve connected components that have a connectivity relationship with the target component. By classifying and processing different types of query needs, corresponding professional algorithms can be invoked in a targeted manner, ensuring both the accuracy of the query results and improving query efficiency.

[0041] Furthermore, step S103 specifically includes the following steps: Step B1: Obtain the target operation interface that matches the control intention from the preset operation interface specifications using the model control tool; The "Operation Interface Specification" refers to the Application Programming Interface (API) document provided by the BIM model management platform (such as BIMFace), which defines various model operation functions that can be called and their parameter formats. Upon receiving a control intent, the model control tool first matches the corresponding target operation interface from the Operation Interface Specification based on the intent type (such as style modification, highlighting, view switching, etc.). In practical applications, the target operation interface corresponding to the standard interaction type in the operation intent can be obtained from the Operation Interface Specification.

[0042] Step B2: Combine the identity information with the target operation interface using the model manipulation tool to generate control code that conforms to the predefined syntax specification; The model manipulation tool combines identification information (such as the component's originalID) with the target operation interface, and assembles it into complete control code according to a predefined syntax specification through a code generation engine. This predefined syntax specification requires that the generated control code must include three basic parts: model object acquisition, operation function call, and view rendering refresh, and the code format must conform to the requirements of the BIM model management platform's execution environment. In practical applications, standard attribute items in the operation intent and / or the standard component type, identification information, and target operation interface can be combined to generate control code.

[0043] Step B3: Execute the control code through the model manipulation tool to perform the specified operation on the target building object; The model manipulation tool directly runs the generated control code by calling the code execution environment of the BIM model management platform to perform specified operations on the target building object.

[0044] It should be noted that the control code is executable JavaScript code, and it is prohibited from containing comments, explanations, Markdown formatting, or other non-code content, and must strictly adhere to modern JavaScript syntax (such as ES6+). In this embodiment, room highlighting (display) can be achieved by specifying a room ID and a specified interface; component highlighting (display) can be achieved by specifying a component ID and a specified interface; and component color can be modified by specifying a component ID and color attribute.

[0045] Furthermore, the method also includes: Based on the query results and / or the operation results of performing the specified operation, generate visual interactive results. The interaction result information includes at least one of the following: text description data, structured data table, and model view corresponding to the control code.

[0046] In this embodiment, based on the query result information (such as component attribute data, collision relationship list, and connected component information) generated under the query intent and / or the feedback information after performing the specified operation under the control intent (such as highlight success prompt and color modification confirmation), visual interactive result information is generated, thereby ensuring that users can quickly understand the query results or control effects and improving the intuitiveness of the interactive experience.

[0047] In this embodiment, based on AI dialogue agents and natural language interaction, a "dialogue-as-operation" mode for BIM review and verification is realized. Through multi-agent collaboration involving business explanation, orchestration execution, and code generation, complex 3D model operations are transformed into intuitive language commands. Components and their attributes that meet user requirements can be directly displayed in tabular form, supporting user selection and positioning, significantly reducing the professional threshold and improving review efficiency. Natural language commands directly drive BIM models to perform 3D navigation, component highlighting, attribute querying, and spatial sectioning; achieving a lightweight "dialogue-as-operation" review workflow, transforming traditional graphical interface operations into language interaction, completely eliminating barriers to 3D software operation, enabling review experts with CAD backgrounds to conduct 3D model reviews with zero barriers; when reviewing in conjunction with clause review results, a query and analysis of the review route is provided, allowing reviewers to further understand the results.

[0048] Example 2 This invention provides a device for intelligently controlling BIM models, such as... Figure 2 As shown, the device specifically includes the following components: The conversion module 201 is used to receive natural language interaction instructions for the target BIM model and convert the natural language interaction instructions into structured instructions including the target building object and the interaction intent using a preset building domain knowledge base. The query module 202 is used to obtain relevant model data about the target building object from the target BIM model through a data query tool if the interaction intent is a query intent, and to form query result information that satisfies the query intent based on the relevant model data; The control module 203 is used to, if the interaction intent is a control intent, first determine the identity information of the target building object from the target BIM model through the data query tool, and then generate and execute control code corresponding to the control intent and the identity information through the model control tool to perform a specified operation on the target building object.

[0049] Specifically, the conversion module 201 is used for: The natural language interaction commands are parsed to extract building object nouns, building object attribute nouns, component nouns, and interaction verbs; Determine whether the architectural object name exists in the business object document in the architectural domain knowledge base. If so, obtain the object explanation information associated with the architectural object name from the business object document, and form the target architectural object based on the architectural object name and the object explanation information. Using the object attribute documents in the aforementioned architectural knowledge base, the architectural object attribute names are converted into standard attribute items; Using the basic component documents in the aforementioned architectural knowledge base, the component names are converted into standard component types; Using the interaction intent documents in the aforementioned architectural knowledge base, the interaction verbs are converted into standard interaction types; The interaction intent is formed based on the standard attribute items, the standard component types, and the standard interaction types.

[0050] Furthermore, the query module 202 is specifically used for: When the standard interaction type in the query intent is component and attribute query, a first component belonging to the target building object and satisfying the standard component type is determined from the target BIM model, and the attribute values ​​of the first component with respect to the standard attribute items are obtained. The query result information is then formed based on the first component and the attribute values; or, When the standard interaction type in the query intent is a collision query, a second component belonging to the target building object and satisfying the standard component type is determined from the target BIM model, and collision components that have a collision relationship with the second component are determined. The query result information is then formed based on the second component and the collision components; or, When the standard interaction type in the query intent is a connectivity query, a third component belonging to the target building object and satisfying the standard component type is determined from the target BIM model, and a connecting component that has a connectivity relationship with the third component is determined. The query result information is then formed based on the third component and the connecting component.

[0051] Furthermore, the control module 203 is specifically used for: The model manipulation tool obtains a target operation interface that matches the manipulation intention from a preset operation interface specification; The model manipulation tool combines the identity information with the target operation interface to generate control code that conforms to a predefined syntax specification. The control code is executed by the model manipulation tool to perform specified operations on the target building object.

[0052] Furthermore, the device also includes a display module for: Based on the query results and / or the operation results of performing the specified operation, generate visual interactive results. The interaction result information includes at least one of the following: text description data, structured data table, and model view corresponding to the control code.

[0053] Example 3 This invention provides a system for intelligent control of BIM models, the system specifically comprising: (1) Business Interpretation Agent: Parses the natural language interaction instructions input by the user, and decomposes the natural language interaction instructions input by the user into standard structured instructions based on the preset dictionary (i.e., the knowledge base of the building domain) and the prompt words created in advance for the Business Interpretation Agent; wherein, the structured instructions include: building object, interaction intent type and interaction requirement parameters.

[0054] The business interpretation agent is used to translate "building-related objects, attributes, and actions" in the user's everyday language into standardized information that can be recognized by subsequent agents. For example, if the natural language interaction command is "What is the ID of the bathroom on the 2nd floor?", the business interpretation agent will expand "2nd floor" to "2F, F2, 2nd floor, 2", expand "bathroom" to "room name containing any of the following keywords: bathroom, WC, toilet, washbasin, wash hands", and map "ID" to "element number".

[0055] In this embodiment, the user submits an interaction command containing specific interaction requirements and a scenario description to the business interpretation agent; the business interpretation agent encapsulates constraints such as prompt words and session history; and finally, the business interpretation agent outputs a structured result to interpret the business name in the interaction command.

[0056] (2) Arrange and execute Agent: According to the interaction intent type, the large language model automatic matching tool calls the strategy, establishes a communication channel with the BIM model data source through MCP (Model Control Protocol), and executes various data query functions; in addition, when the interaction intent type involves model control, the code generation Agent is called, and control code that can be executed by BIMFace is dynamically generated based on the interaction requirement parameters in the structured instructions, so as to realize in-depth control of BIMFace.

[0057] The processing logic for orchestrating and executing the Agent is as follows: First, it determines whether the user's interaction request is a BIM model operation request (i.e., manipulation intent, such as model component editing, attribute adjustment, etc.). If so, the data query class MCP tool (i.e., data query tool) must be called first to obtain the component's originalID (i.e., identification information) and component type required for the operation. After the MCP tool returns the data, the code-generated Agent is called to execute the model control operation. In addition, if the required component is not found, there is no need to call the code-generated Agent. If not, if the user needs to query whether the component exists, the number of queries, component attributes, collisions, connectivity, etc., the data query class MCP tool must be called first. Subsequently, it is necessary to determine whether other Agents need to be called. After obtaining the specified information through the MCP tool or other Agents, the system will return the obtained content and perform subsequent processing until a final answer can be generated. In addition, if the existing session records can directly yield the answer, the answer will be given directly.

[0058] In this embodiment, when the interaction intent type is a query intent, a model data query function is built based on the GFC standard and MCP protocol to realize the source of model data during Agent usage. Currently, three query methods based on the GFC standard are supported: component and attribute query, which queries and returns a structured list of required components and component attribute values ​​based on the query statement; collision query, which executes the component collision method based on the query statement, obtains a list of components that the required component has collided with, and returns it in a structured manner; and logical relationship query, which queries the connectivity relationships of components such as pipes and ducts based on the query statement and returns the results in a structured manner. In addition, the above three tool-defined query DSL syntaxes provide complex query capabilities, thereby significantly reducing the excessive tokens generated by tool calls and supporting multi-condition filtering and attribute matching judgments.

[0059] In this embodiment, the orchestration execution agent encapsulates operations such as prompt words and session history; the orchestration execution agent determines whether to call a data query tool or code generation agent based on structured instructions; the output results of the orchestration execution agent include: a text field for directly displaying the echoed text, a table field for displaying the list of components involved in the interactive instructions, and a code field for displaying the control code for operating BIMFace.

[0060] (3) Code generation agent: Generate corresponding control code based on the BIM model control requirements proposed by the orchestration execution agent and the operation API in the prompt words generated in advance for the code generation agent.

[0061] In this embodiment, a multi-agent collaborative structure is designed based on the LLM large language model. A three-layer agent mechanism is used to realize the full-process conversion from natural language interaction commands to BIMFace model operations. The three agents work together. After the user's question is handled by the business interpretation agent, the generated result is passed to the orchestration execution agent. During the operation of the orchestration execution agent, the large language model judges and generates the result of calling the MCP tool and code generation agent. After the orchestration execution agent executes the result, it returns the result to the large language model for further processing to form the final interaction result.

[0062] In addition, in this embodiment, the Spring AI Advisor interface can be used to embed dialogue tracking capabilities throughout the entire process, so as to record the input, output and execution time of each Agent and various tool calls, and monitor the response time in real time so as to call alarms when timeouts occur; the ChatMemory interface can also be used to maintain the dialogue context and support long session review tasks.

[0063] Example 4 This embodiment also provides a computer device, such as a smartphone, tablet computer, laptop computer, desktop computer, rack server, blade server, tower server, or cabinet server (including a standalone server or a server cluster composed of multiple servers), etc., capable of executing programs. Figure 3 As shown, the computer device 30 in this embodiment includes, but is not limited to, a memory 301 and a processor 302 that are communicatively connected to each other via a system bus. It should be noted that... Figure 3 Only a computer device 30 with components 301-302 is shown; however, it should be understood that it is not required to implement all of the components shown, and more or fewer components may be implemented instead.

[0064] In this embodiment, the memory 301 (i.e., the readable storage medium) includes flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 301 may be an internal storage unit of the computer device 30, such as the hard disk or memory of the computer device 30. In other embodiments, the memory 301 may also be an external storage device of the computer device 30, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the computer device 30. Of course, the memory 301 may include both the internal storage unit and the external storage device of the computer device 30. In this embodiment, the memory 301 is typically used to store the operating system and various application software installed on the computer device 30. In addition, the memory 301 may also be used to temporarily store various types of data that have been output or will be output.

[0065] In some embodiments, processor 302 may be a central processing unit (CPU), controller, microcontroller, microprocessor, or other data processing chip. This processor 302 is typically used to control the overall operation of computer device 30.

[0066] Specifically, in this embodiment, the processor 302 is used to execute a program for a method of intelligently controlling a BIM model stored in the memory 301. When the program for the method of intelligently controlling a BIM model is executed, it performs the following steps: Receive natural language interaction instructions for the target BIM model, and use a preset building domain knowledge base to convert the natural language interaction instructions into structured instructions that include the target building object and the interaction intent; If the interaction intent is a query intent, then relevant model data about the target building object is obtained from the target BIM model through the data query tool, and query result information that satisfies the query intent is formed based on the relevant model data; If the interaction intent is a control intent, the identification information of the target building object is first determined from the target BIM model by the data query tool, and then the control code corresponding to the control intent and the identification information is generated and executed by the model control tool to perform the specified operation on the target building object.

[0067] For a detailed description of the above method steps, please refer to Example 1. This example will not be repeated here.

[0068] Example 5 This embodiment also provides a computer-readable storage medium, such as flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, server, app store, etc., which stores a computer program. When the computer program is executed by a processor, it implements the following method steps: Receive natural language interaction instructions for the target BIM model, and use a preset building domain knowledge base to convert the natural language interaction instructions into structured instructions that include the target building object and the interaction intent; If the interaction intent is a query intent, then relevant model data about the target building object is obtained from the target BIM model through the data query tool, and query result information that satisfies the query intent is formed based on the relevant model data; If the interaction intent is a control intent, the identification information of the target building object is first determined from the target BIM model by the data query tool, and then the control code corresponding to the control intent and the identification information is generated and executed by the model control tool to perform the specified operation on the target building object.

[0069] For a detailed description of the above method steps, please refer to the first embodiment. This embodiment will not repeat the details here.

[0070] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0071] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0072] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method.

[0073] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A method for intelligent control of a BIM model, characterized in that, The method includes: Receive natural language interaction instructions for the target BIM model, and use a preset building domain knowledge base to convert the natural language interaction instructions into structured instructions that include the target building object and the interaction intent; If the interaction intent is a query intent, then relevant model data about the target building object is obtained from the target BIM model through the data query tool, and query result information that satisfies the query intent is formed based on the relevant model data; If the interaction intent is a control intent, the identification information of the target building object is first determined from the target BIM model by the data query tool, and then the control code corresponding to the control intent and the identification information is generated and executed by the model control tool to perform the specified operation on the target building object.

2. The method for intelligent control of BIM models according to claim 1, characterized in that, The process of converting the natural language interaction instructions into structured instructions, including the target building object and the interaction intent, using a pre-set architectural domain knowledge base includes: The natural language interaction commands are parsed to extract building object nouns, building object attribute nouns, component nouns, and interaction verbs; Determine whether the architectural object name exists in the business object document in the architectural domain knowledge base. If so, obtain the object explanation information associated with the architectural object name from the business object document, and form the target architectural object based on the architectural object name and the object explanation information. Using the object attribute documents in the aforementioned architectural knowledge base, the architectural object attribute names are converted into standard attribute items; Using the basic component documents in the aforementioned architectural knowledge base, the component names are converted into standard component types; Using the interaction intent documents in the aforementioned architectural knowledge base, the interaction verbs are converted into standard interaction types; The interaction intent is formed based on the standard attribute items, the standard component types, and the standard interaction types.

3. The method for intelligent control of BIM models according to claim 2, characterized in that, If the interaction intent is a query intent, then relevant model data about the target building object is obtained from the target BIM model through a data query tool, and query result information that satisfies the query intent is formed based on the relevant model data, including: When the standard interaction type in the query intent is component and attribute query, a first component belonging to the target building object and satisfying the standard component type is determined from the target BIM model, and the attribute values ​​of the first component with respect to the standard attribute items are obtained. The query result information is then formed based on the first component and the attribute values; or, When the standard interaction type in the query intent is a collision query, a second component belonging to the target building object and satisfying the standard component type is determined from the target BIM model, and collision components that have a collision relationship with the second component are determined. The query result information is then formed based on the second component and the collision components; or, When the standard interaction type in the query intent is a connectivity query, a third component belonging to the target building object and satisfying the standard component type is determined from the target BIM model, and a connecting component that has a connectivity relationship with the third component is determined. The query result information is then formed based on the third component and the connecting component.

4. The method for intelligent control of BIM models according to claim 1, characterized in that, The step of generating and executing control code corresponding to the control intent and the identity information through a model manipulation tool to perform a specified operation on the target building object includes: The model manipulation tool obtains a target operation interface that matches the manipulation intention from a preset operation interface specification; The model manipulation tool combines the identity information with the target operation interface to generate control code that conforms to a predefined syntax specification. The control code is executed by the model manipulation tool to perform specified operations on the target building object.

5. The method for intelligent control of BIM models according to claim 1, characterized in that, The method further includes: Based on the query results and / or the operation results of performing the specified operation, generate visual interactive results. The interaction result information includes at least one of the following: text description data, structured data table, and model view corresponding to the control code.

6. A device for intelligent control of BIM models, characterized in that, The device includes: The conversion module is used to receive natural language interaction instructions for the target BIM model and convert the natural language interaction instructions into structured instructions including the target building object and the interaction intent using a preset building domain knowledge base. The query module is used to obtain relevant model data about the target building object from the target BIM model through a data query tool if the interaction intent is a query intent, and to form query result information that satisfies the query intent based on the relevant model data; The control module is used to, if the interaction intent is a control intent, first determine the identity information of the target building object from the target BIM model through the data query tool, and then generate and execute control code corresponding to the control intent and the identity information through the model control tool to perform a specified operation on the target building object.

7. The device for intelligent control of BIM models according to claim 6, characterized in that, The control module is used for: The model manipulation tool obtains a target operation interface that matches the manipulation intention from a preset operation interface specification; The model manipulation tool combines the identity information with the target operation interface to generate control code that conforms to a predefined syntax specification. The control code is executed by the model manipulation tool to perform specified operations on the target building object.

8. The device for intelligent control of BIM models according to claim 6, characterized in that, The device further includes a display module for: Based on the query results and / or the operation results of performing the specified operation, generate visual interactive results. The interaction result information includes at least one of the following: text description data, structured data table, and model view corresponding to the control code.

9. A computer device, the computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor, when executing the computer program, implements the steps of the method according to any one of claims 1 to 5.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images