WO2021114479A1

WO2021114479A1 - Three-dimensional display system and method for sound control building information model

Info

Publication number: WO2021114479A1
Application number: PCT/CN2020/076321
Authority: WO
Inventors: 林佳瑞
Original assignee: 清华大学
Priority date: 2019-12-11
Filing date: 2020-02-24
Publication date: 2021-06-17
Also published as: CN110889161A; CN110889161B

Abstract

A three-dimensional display system and method for a sound control building information model. The system comprises: a voice input module, a voice recognition and processing module, a command processing module and a three-dimensional display module, wherein the voice input module is used for collecting a user voice and transmitting the user voice to the voice recognition and processing module in a standard voice data format; the voice recognition and processing module converts the user voice into a command that can be recognized by a BIM and transmits the command to the command processing module; the command processing module parses and executes the received command and transmits an execution result of the command to the three-dimensional display module; and according to the execution result of the received command, the three-dimensional display module updates a three-dimensional display effect of the BIM and outputs same in real time. The present invention realizes the automatic processing conversion of a user voice input in a BIM and the generation, execution and automatic update of a three-dimensional display command, and effectively improves the efficiency of a user interactively controlling the three-dimensional display of BIM data.

Description

Three-dimensional display system and method of voice-controlled building information model

Technical field

The invention relates to a three-dimensional display system and method for a voice-controlled building information model, and belongs to the technical field of building information.

Background technique

Building Information Model (BIM) is an engineering data model based on three-dimensional digital technology that integrates various related information of construction projects, and is a digital expression of the physical and functional characteristics of engineering project facilities. A complete building information model can connect data, processes, and engineering resources at different stages of the life of a building project, and a complete description of the engineering object, which can be commonly used by construction project owners, designers, constructors, and users. BIM is a unified engineering data source for all stages of construction projects, which fully integrates various data such as text, pictures, and monitoring designed at each stage of engineering design, construction and operation, and supports all parties to query, utilize, update and store in the process of work Information to assist project management and decision-making.

Because construction projects have the characteristics of large volume, long period, and many specialties, a large amount of data is used and generated in each stage of design, construction and operation. At the same time, it also involves various professional fields such as schedule, cost, and quality. Each stage and field often Different software and data formats are adopted, and various stages and fields are interdependent. Therefore, construction project data has typical big data characteristics. In addition, each professional business of construction engineering relies on other professional data. The project management and operation stage involves the comprehensive analysis and utilization of multiple professional data. At the same time, it also requires relevant engineering personnel to master multiple professional knowledge, and put forward proposals for the generation, management and use of engineering data. High demands.

BIM includes complete design data for various disciplines such as architecture, structure, electromechanical pipeline, etc., including the three-dimensional shape of each component of the building, as well as the characteristics and parameters of the corresponding shape. In order to achieve intuitive and quick interaction between project owners, designers, construction parties and users, the three-dimensional display or visualization of BIM is essential. Existing BIM software all rely on a 3D display engine for 3D model display and interaction. In order to enhance the realism of BIM, it is often necessary to use model rendering, baking and other technologies to improve the effect. At the same time, in order to combine the needs of different scenarios, it is often necessary to use different hardware terminals such as desktop computers, virtual reality helmets, augmented reality and mixed reality glasses to perform the three-dimensional display and visualization of BIM. Therefore, how to efficiently and quickly control the 3D display and visualization effects of BIM is very important.

Currently, there are mainly four BIM three-dimensional display and visualization control and interaction methods used in desktop computers, head-mounted displays, tablet computers, augmented reality and mixed reality glasses and other devices:

(1) Three-dimensional display interactive control method based on mouse and keyboard

This method is mainly used for BIM software running on desktop computer equipment. The 3D display engine of the BIM software often implements interactive operations such as rotation, zooming, and translation of the BIM 3D display through operations such as mouse clicks, drags, and scroll wheels. At the same time, the first-person perspective of the three-dimensional scene roaming can also be realized through keyboard commands. In addition, the selection, display and hiding of professional components in BIM can be selected by mouse clicking, frame selection of corresponding components, and corresponding functions can be realized based on clicking menu commands or keyboard shortcut commands. This method requires the user to have a good understanding of BIM software and know which interactive method to use to realize the 3D display control of the model.

(2) Three-dimensional display interactive control method based on touch screen

This method is similar to method (1), and is mainly used for devices such as desktop computers, tablet computers, or mobile phones with touch screens. This type of equipment can simulate mouse clicks, scroll wheels and other operations through interactive methods such as finger click, two-finger opening and closing, and dragging, so as to realize the process of selecting, zooming, and panning various professional components in BIM. At the same time, the display and hiding of the BIM model and the interaction of the three-dimensional scene roaming can also be realized through the touch screen clicking menu commands.

(3) Three-dimensional display interactive control method based on handle or dedicated input device

This method is mainly used for head-mounted display devices such as virtual reality helmets. Users can interact with the BIM three-dimensional display scene by means of virtual reality handles and other methods. For example, the cursor can be moved in the virtual scene by moving the handle, and the cursor can be clicked by clicking the handle button. Therefore, the corresponding operation can be equivalent to the way (1) Mouse movement, click and other operations, so as to realize the selection of BIM model, the click of menu commands and other functions, and the corresponding interactive control of BIM model selection, display, and hiding can also be realized. .

(4) Three-dimensional display interactive control method based on gestures and body posture

This method is mainly used for devices such as augmented reality eyes, mixed reality eyes, and virtual reality helmets. First, the user can simulate the mouse click operation in the similar way (1) by operating a virtual point in the air to realize BIM model selection or menu command click, so as to realize the interaction of BIM model display and hiding; second, the user can wave the palm of the hand , The arm realizes operations such as translation or rotation of the model, and the zooming interaction of the model can also be realized by opening and closing the hands or pinching the palm; finally, the user can use the change of the body posture to realize the interaction of the BIM three-dimensional display, such as by turning, moving forward and backward The roaming function of the first person perspective in the virtual scene.

With the continuous popularization of BIM technology, the scale of BIM-based construction engineering data has become increasingly large, and the number of BIM users has also grown rapidly. How to provide users with more intuitive and convenient three-dimensional display interactive control, efficient use of BIM data for users, and multi-party implementation based on BIM Collaboration is very important. However, the display method in the prior art described above still has problems such as unfriendly interaction, complicated commands or operations, and numerous menu options. The specific analysis is as follows:

(1) The model display and hide operation is cumbersome: If you need to show or hide some components in BIM, first, you need to select the components by mouse click/frame selection, finger click, etc., then you need to switch commands by clicking the menu, and finally click The display or hide command realizes the function of model display and hide control. If there are more model components and more software menu commands, this process often requires multiple clicks and menu command switching to complete. However, BIM often contains multiple professional data, and the typical user scenario usually only needs to view the model of each single professional or area, which means that it is often necessary to select many components, and control the display and hiding of each component. At the same time, the existing BIM software is mainly design and construction management software. Its interface and functions involve multiple majors, with multiple functions and complex interface layout. It takes a long time for relevant engineers to find or click on relevant menu commands. Therefore, for BIM software users, the model display and hide control operations are often cumbersome, and this function is sometimes used frequently, so it will take up a lot of user time, seriously affecting the efficiency of viewing and using BIM data, and restricting the collaboration of all participants. jobs. The interactive mode of gesture or body posture is often not suitable for the display and hide control of the model.

(2) View and viewpoint switching operations are complicated: In the process of BIM 3D roaming and model viewing, users often create or record different model viewing angles and viewpoints, and also create east facade, south facade, and a certain level of planes, etc. Different views. However, when switching views, users often need to open the corresponding view or viewpoint list by clicking a menu command, and then select and switch to the corresponding view or viewpoint. When there are many viewpoints or views, users will spend a lot of time searching and searching for data. Therefore, comprehensive consideration of the process of searching and clicking menu commands and searching for views and viewpoints will also take up a lot of users' time and affect the efficiency of users' work and collaboration. The way of using gesture interaction is similar to the way of clicking, and the body posture is basically not suitable for switching between views and viewpoints.

(3) Highly dependent on palm and arm movement: From the above two, it can be seen that the current interactive methods provided by relevant BIM software and equipment are highly dependent on interactive methods such as mouse clicks, keyboard commands or gestures. These interactive methods require repeated movement of the palm or arm. Users are very prone to fatigue. Therefore, the existing three-dimensional display interactive control mode is difficult to adapt to working scenarios such as long-term BIM data viewing and use, and it will bring greater adverse effects on the health of users.

Summary of the invention

In view of the above problems, the purpose of the present invention is to provide a voice-activated building information model three-dimensional display system and method, which realizes the automatic processing and conversion of user voice input, the generation and execution of three-dimensional display commands, and the automatic update of three-dimensional display effects, which effectively improves the design Teachers, managers, owners and other users interactively control the efficiency of the three-dimensional display of BIM data, which reduces learning costs and improves the user-friendliness of BIM software.

The invention discloses a three-dimensional display system for a voice-controlled building information model, which includes: a voice input module, a voice recognition and processing module, a command processing module, and a three-dimensional display module; the voice input module is used to collect user voice and convert the user voice to standard voice The data format is transmitted to the voice recognition and processing module; the voice recognition and processing module converts the user's voice collected by the voice input module into commands that can be recognized by the building information model, and transmits the commands to the command processing module; the command processing module performs commands on the received commands Analyze and execute, and transmit the execution result of the command to the 3D display module; the 3D display module updates and outputs the 3D display effect of the building information model in real time according to the execution result of the received command.

Further, the voice recognition and processing module includes a three-dimensional display command database, and the three-dimensional display command database stores three-dimensional display commands related to the building information model and the corresponding operation objects of the commands.

Further, the speech recognition and processing module includes a building information model metadata database, a metadata model for storing building information model data in the building information model metadata database, and the correspondence between the metadata of the building information model and the three-dimensional display commands in the three-dimensional display command database. Metadata The model includes the data entities and their types and attributes contained in the building information model.

Furthermore, the voice recognition and processing module is a general machine learning or deep learning voice recognition model, which realizes the conversion of user voice input to text, and combines the metadata of the building information model and the three-dimensional display command data to process the text generated by the voice input; The latter text is grammatically analyzed, and the generated grammar tree is preliminarily checked to ensure that the user's voice input generates effective three-dimensional display control and interactive commands.

Further, the command processing module includes a command parsing module and a command execution module. The command parsing module is based on the output commands of the voice recognition and processing module, and generates building information model data or views and viewpoints based on the relationship between the three-dimensional display commands and their operating objects At the same time, a three-dimensional display command call sequence is generated, and the filter command sequence and the three-dimensional display command call sequence are constructed as the data processing and command call command stream of the building information model, and the command stream is transmitted to the command execution module.

Further, the command execution module executes the command stream output by the command analysis module, obtains and queries data related to the command stream, applies the command stream to the relevant data, and sends the command execution result to the three-dimensional display module.

The present invention also provides a voice-activated building information model three-dimensional display method, which is implemented by any of the above-mentioned voice-activated building information model three-dimensional display systems, including the following steps: S1 acquires voice input in real time, and converts the voice input into a standard voice data format for transmission Go to the next step; S2 converts the user's voice collected by the voice input module into commands that can be recognized by the building information model, and transmits the commands to the command processing module; S3 command processing module parses and executes the received commands to generate building information model data Or view, viewpoint filtering command sequence and three-dimensional display command call sequence, construct the data processing and command call command flow of building information model; execute data processing and command call command flow, automatically obtain and query data related to the command flow, and send commands The flow acts on the relevant data and sends the command execution result to the 3D display module; S4 automatically updates the display result of the 3D display module based on the command execution result, realizing dynamic interaction with the user and presentation of the display result.

Further, S2 specifically includes the following steps: S2.1 uses algorithm modules such as machine learning or deep learning to realize the automatic conversion of user voice input to text, and generates text input corresponding to the voice input; S2.2 uses statistical models and machine learning models Or deep learning model to segment the text; S2.3 uses natural speech processing models and algorithms to tag the word segmentation results; S2.4 text disambiguation: clean up stop words and unify synonyms into a standard word; S2 .5 Based on the part-of-speech tagging in step S2.3 and the result of text disambiguation in step S2.4, a natural speech processing algorithm is used to construct a grammar tree of the user input text.

Furthermore, S2 also includes: S2.6 importing the data in the 3D display command database and the building information model metadata database, adjusting the word segmentation result in step S2.2 according to the data in the database, and combining the word segmentation result and the grammar tree construction result with the read database Compare and analyze the data in the middle, check and confirm that all parts of the grammar tree can be converted into 3D display control commands and their operation object information; if they cannot be converted into 3D display control commands, the voice control process will be terminated and the next user voice input will be awaited.

Further, step S3 specifically includes the following steps: S3.1 receives the command generated in the voice input module, and imports the data in the three-dimensional display command database and the building information model metadata database; S3.2 identifies and determines which building information model data needs to be extracted Or view, viewpoint; S3.3 generates a data object filtering command sequence corresponding to the user's voice command according to the recognition result in step S3.2; S3.4 recognizes and determines the three-dimensional display command information that needs to be executed; S3.5 according to step S3. 4, the three-dimensional display command recognition result is generated, and the three-dimensional display command call sequence corresponding to the user's voice command is generated; S3.6 sorts the generated data object filtering command sequence and the three-dimensional display command sequence to ensure related commands and execution order, forming data processing and The command invokes the command flow.

The present invention has the following advantages due to the above technical scheme: The system and method in this article realize the automatic processing and conversion of user voice input, the generation and execution of three-dimensional display commands, and the automatic update of three-dimensional display effects, which effectively improves designers and management personnel. , Owners and other users interactively control the efficiency of the three-dimensional display of BIM data, which reduces learning costs and improves the user-friendliness of BIM software. The system and method in this article can support a variety of BIM three-dimensional display interactive control based on voice recognition, and can use different voice input terminals, different voice recognition algorithm modules, different BIM metadata and three-dimensional display command storage methods, and support different BIM software , 3D display module and virtual reality, augmented reality, and mixed reality environments provide a method for owners, designers, managers, etc. to efficiently and interactively control the BIM 3D display effect.

Description of the drawings

FIG. 1 is a schematic structural diagram of a three-dimensional display system of a voice-controlled building information model in an embodiment of the present invention;

FIG. 2 is a flowchart of a three-dimensional display method of a voice-controlled building information model in an embodiment of the present invention;

3 is a flowchart of a working method of a speech recognition and processing module in an embodiment of the present invention;

4 is a flowchart of a working method of a command parsing module in an embodiment of the present invention;

FIG. 5 is a flowchart of the working method of the command execution module and the three-dimensional display module in an embodiment of the present invention.

Detailed ways

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings. However, it should be understood that the drawings are only provided for a better understanding of the present invention, and they should not be construed as limiting the present invention. In the description of the present invention, it should be understood that the terms are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance.

Example one

This embodiment discloses a three-dimensional display system of a voice-controlled building information model, as shown in FIG. 1, which includes:

(1) Voice input module: This module is mainly a voice input device such as a microphone of a computer device such as a desktop computer, a tablet computer, and a head-mounted display, but it can also be a voice input device other than a microphone. The voice input module can collect user voice input in real time in different environments, convert the collected voice input into a standard voice data format, and transmit the voice input to the voice recognition and processing module in the standard voice data format.

(2) Speech recognition and processing module: Based on general machine learning or deep learning speech recognition models, this module converts user voice input into text form, and combines BIM metadata and three-dimensional display command data to process the text generated by voice input . This module will perform grammatical analysis on the processed text and conduct a preliminary check on the generated grammar tree to ensure that the user's voice input generates effective three-dimensional display control and interactive commands. This module can further optimize the results of text processing and grammatical analysis by continuously learning the user's voice habits, so that users can achieve BIM operations without using professional terminology.

(3) Command processing module: The BIM metadata and 3D display command data in this module parse and execute the commands received from the voice recognition and processing module, and generate building information model data or views, filtering command sequences of viewpoints, and 3D display Command call sequence, build the data processing and command call command flow of building information model; execute data processing and command call command flow, automatically obtain and query data related to the command flow, and apply the command flow to related data, and send the command execution result To the three-dimensional display module.

(4) Three-dimensional display module: This module is mainly the display screen of different computer equipment, virtual reality helmet display and other projection and display equipment, etc. This module will be based on the command execution result of the command processing module and the display update or transformation operation on the model. The 3D display effect is updated and output in real time, so as to realize the dynamic interactive 3D display control with the user.

Among them, the voice recognition and processing module includes a three-dimensional display command database and a BIM metadata database: the three-dimensional display command database is stored in a graph database or an ontology database, and the three-dimensional display commands related to BIM software and the corresponding operation objects of the commands. The operation objects include the views corresponding to the commands. Name, viewpoint name, or component category, etc. On the one hand, the 3D display command database provides support for the voice recognition and processing module to ensure that the user’s voice input can be converted into the corresponding 3D display command or the operating object of the command. On the other hand, it also analyzes and converts the subsequent commands to the 3D display commands of the BIM software. Sequence provides support. The three-dimensional display command database can also adopt other database forms such as relational database, document database, column database and so on. The BIM metadata database mainly adopts the form of graph database or ontology database, which stores the metadata model of the corresponding BIM data of BIM software, that is, the data entities contained in BIM and their types, attributes and other information, as well as the BIM metadata and three-dimensional display command database Correspondence of three-dimensional display commands. On the one hand, the BIM metadata database provides support for speech recognition and processing modules to ensure that the user’s voice input correctly converts BIM data objects and their corresponding three-dimensional display commands. On the other hand, it also lays the foundation for subsequent command analysis and conversion into the three-dimensional display command sequence of BIM software. basis. The BIM metadata database can also adopt other database forms such as relational database, document database, column database and so on.

In this embodiment, the command processing module includes a command parsing module and a command execution module. The command parsing module is based on the output commands of the voice recognition and processing module, combined with the three-dimensional display command database and the BIM metadata database, according to the three-dimensional display commands and their operation objects The relationship between the building information model data or the view and viewpoint filtering command sequence is generated, and the 3D display command calling sequence is generated at the same time, and the filtering command sequence and the 3D display command calling sequence are constructed as the data processing and command calling command flow of the building information model , And transmit the command stream to the command execution module. The command execution module executes the command stream output by the command analysis module, obtains and queries data related to the command stream, applies the command stream to the relevant data, and sends the command execution result to the three-dimensional display module.

Example two

Based on the same inventive concept, this embodiment discloses a three-dimensional display method of a voice-activated building information model, as shown in Fig. 2-5, including the following steps:

S1 uses the voice input module to obtain voice input in real time, and convert the voice input into a standard voice data format for transmission to the next step;

S2 converts the user's voice collected by the voice input module into commands that can be recognized by the building information model, and transmits the commands to the command processing module; uses the three-dimensional display command database and the three-dimensional display commands and BIM metadata stored in the BIM metadata database to input the voice into the module The output standard speech format data is converted into text, and the text segmentation, stop word clarity, and synonym disambiguation are processed, and then the corresponding grammar tree is constructed, and the grammar tree is checked to determine whether a valid command sequence can be generated. If a valid user command sequence cannot be generated, output a prompt to the user and wait for the next user voice input. If a valid user command sequence can be generated, then enter the user command parsing step.

The S3 command processing module parses and executes the received commands, generates the building information model data or view, the filtering command sequence of the viewpoint, and the three-dimensional display command call sequence, constructs the data processing and command call command flow of the building information model; performs data processing and Command calls the command stream, automatically obtains and queries data related to the command stream, applies the command stream to the relevant data, and sends the command execution result to the three-dimensional display module;

S4 automatically updates the display result of the 3D display module based on the command execution result, realizing dynamic interaction with the user and the presentation of the display result;

The S5 command ends and waits for the next user voice input.

The purpose of the data input processing and conversion process is to convert the user's voice input into text and perform subsequent processing to ensure that the user input can generate an effective three-dimensional display control command sequence. As shown in Figure 3, the main implementation methods of this step are as follows:

S2.1 Speech-to-text: Using algorithm modules such as machine learning or deep learning to realize the automatic conversion of user's voice input to text, and generate text input corresponding to the voice input;

S2.2 Text segmentation: Use statistical models, machine learning models or deep learning models to segment texts;

S2.3 Part-of-speech tagging: Use natural speech processing models and algorithms to tag the word segmentation results, including verbs, nouns, quantifiers, etc.;

S2.4 Text disambiguation: clean up stop words based on stop word dictionaries, synonym dictionaries and other natural speech processing methods, and unify synonyms into standard forms, laying the foundation for subsequent analysis;

S2.5 grammar tree construction: based on the results of part-of-speech tagging in step S2.3 and text disambiguation in step S2.4, a natural speech processing algorithm is used to construct a grammar tree of user input text;

S2.6 Import the data in the three-dimensional display command database and the building information model metadata database, adjust the word segmentation result in step S2.2 according to the data in the database, compare the word segmentation result and the grammar tree construction result with the data in the read database, and check it And confirm that all parts of the grammar tree can be converted into 3D display control commands and their operation object information; if they cannot be converted into 3D display control commands, the voice control process will be terminated and the next user voice input will be awaited.

The function of the user command parsing link is to combine the results of text segmentation, part-of-speech tagging and grammar tree output in step S2 to generate corresponding data object extraction commands and three-dimensional display control command sequences in combination with the characteristics of the BIM software. As shown in Figure 4, the main implementation method of this step is as follows:

S3.1.1 Support data reading: Read in the data of the three-dimensional display command database and the BIM metadata database as supporting data for the subsequent user command parsing process;

S3.1.2 Command operation object recognition: According to the results of text segmentation and part-of-speech tagging, combined with the read BIM metadata and three-dimensional display command data, identify and determine which BIM data or view and viewpoint data objects need to be extracted;

S3.1.3 Generate object filtering command sequence: According to the recognition result of the command operation object, combined with the data extraction method of the corresponding data object of the BIM software, generate the data object filtering command sequence corresponding to the user's voice command;

S3.1.4 Display command recognition: According to the results of text segmentation and part-of-speech tagging, combined with three-dimensional display command data, identify and determine the three-dimensional display command information that needs to be executed;

S3.1.5 Generate a display command sequence: According to the recognition result of the three-dimensional display command, combined with the implementation method and information of the corresponding command of the BIM software, generate the three-dimensional display command sequence corresponding to the user's voice command;

S3.1.6 Ordering sequence of commands: According to the constructed syntax tree, combined with the dependency of each node of the syntax tree, sort the generated data object filtering command sequence and three-dimensional display command sequence to ensure the corresponding command and order execution.

The main function of user command execution and 3D display update is to update the display results of the 3D display module in real time. As shown in Figure 5, the main implementation method of this step is as follows:

S3.2.1 Execute the object filtering command: execute the data object filtering command generated in step S3.1.3 in the BIM software, and output the filtered data object list;

S3.2.2 Extract command operation object: group and classify the filtered data objects to form the extracted command operation object;

S3.2.3 Execute 3D display command: execute the corresponding 3D display command for the command operation object;

S3.2.4 Send a 3D display update command to the 3D display module to trigger the effect update of the 3D display module;

S3.2.5 Update the 3D display effect of the device: The 3D display module refreshes the display content according to the 3D display update command sent in step S3.2.4, and completes the execution of the user's voice command.

Example three

The following is a comparison of the actual operation method of the present invention on the BIM software Autodesk Revit and the desktop computer

The systems and methods in the first and second embodiments are further described.

(1) Voice input module: This module uses the microphone of the desktop computer as the voice input module, and directly reads the voice commands input by the user in real time;

(2) Speech recognition and processing module: This module and method first use the built-in speech recognition module of the Windows system to realize the speech-to-text function, or the deep learning speech-to-text function realized by TensorFlow or pytorch; the subsequent text segmentation uses the python module jieba To achieve, stop words and synonyms are processed by self-editing algorithm, and syntax tree analysis is implemented by Stanford NLP parser of Stanford University;

(3) Three-dimensional display command database: This module uses the graph database neo4j to store three-dimensional display commands. The main storage content includes Revit display, hide components, and open and switch views. It also includes the data object type and attributes corresponding to the corresponding command. And other information; this module can also be implemented using relational databases, column databases, document databases, etc.;

(4) BIM meta-database: The graph database neo 4j is also used to carry out BIM data related data types, attributes and their inheritance, and association relationships in Revit, in order to construct the correspondence between BIM data objects and text generated by user voice input Provide support; the database can also be implemented using relational databases, column databases, document databases, etc.;

(5) User command analysis module: This module and method adopts C# voice to carry out the secondary development of Revit, and through the corresponding API and voice input module, voice recognition and processing module, three-dimensional display command database, BIM metadata database for mutual call and data Exchange, and finally generate filtering commands and 3D display control command sequences for Revit views, viewpoint elements and component elements;

(6) Command execution module: The module and method are also implemented in the secondary development of Revit using C# voice, so as to realize the filtering and command execution of the corresponding Revit data, and call the command of Revit 3D display update;

(7) Three-dimensional display module: This module is directly composed of the display of the desktop computer and the three-dimensional display interface of Revit. When the user calls the Revit three-dimensional display update command, the three-dimensional display effect of the display can be automatically refreshed, so as to realize the three-dimensional display based on the user's voice control. Display interactive control and update.

The foregoing embodiments are only used to illustrate the present invention. The structure, connection mode, and manufacturing process of each component can be changed. Any equivalent transformation and improvement based on the technical solution of the present invention should not be used. Excluded from the protection scope of the present invention.

Claims

A voice-controlled building information model three-dimensional display system, which is characterized by comprising: a voice input module, a voice recognition and processing module, a command processing module, and a three-dimensional display module;

The voice input module is used to collect user voice, and transmit the user voice to the voice recognition and processing module in a standard voice data format;

The voice recognition and processing module converts the user voice collected by the voice input module into a command that can be recognized by the building information model, and transmits the command to the command processing module;

The command processing module parses and executes the received command, and transmits the execution result of the command to the three-dimensional display module;

The three-dimensional display module updates and outputs the three-dimensional display effect of the building information model in real time according to the execution result of the received command.
The voice-controlled building information model three-dimensional display system according to claim 1, wherein the voice recognition and processing module includes a three-dimensional display command database, and the three-dimensional display command database stores three-dimensional display commands and command responses related to the building information model. The object of operation.
The voice-controlled building information model three-dimensional display system according to claim 2, wherein the voice recognition and processing module includes a building information model metadata database, and the building information model metadata database stores metadata models of building information model data, And the correspondence between the metadata of the building information model and the three-dimensional display command in the three-dimensional display command database, the metadata model including the data entities contained in the building information model and their types and attributes.
The voice-controlled building information model three-dimensional display system according to claim 3, wherein the voice recognition and processing module is a general-purpose machine learning or deep learning voice recognition model, which realizes the conversion of user voice input to text, and combines building information Model metadata and three-dimensional display command data process the text generated by voice input; perform grammatical analysis on the processed text, and perform a preliminary check on the generated syntax tree to ensure that the user’s voice input generates effective three-dimensional display control and interactive commands .
The voice-controlled building information model three-dimensional display system according to any one of claims 2-4, wherein the command processing module includes a command parsing module and a command execution module, and the command parsing module uses the voice recognition and processing Based on the output commands of the module, according to the relationship between the three-dimensional display commands and their operating objects, a sequence of filtering commands for building information model data or views and viewpoints is generated, and a sequence of three-dimensional display commands is generated at the same time. The three-dimensional display command call sequence is constructed as a data processing and command call command stream of the building information model, and the command stream is transmitted to the command execution module.
The voice-controlled building information model three-dimensional display system according to claim 5, wherein the command execution module executes the command stream output by the command analysis module, obtains and queries data related to the command stream, and The command stream is applied to the related data, and the command execution result is sent to the three-dimensional display module.
A method for three-dimensional display of a voice-activated building information model, characterized in that it is implemented by the three-dimensional display system of a voice-activated building information model according to any one of claims 1-6, and comprises the following steps:

S1 obtains voice input in real time, converts the voice input into a standard voice data format and transmits it to the next step;

S2 converts the user voice collected by the voice input module into a command that can be recognized by the building information model, and transmits the command to the command processing module;

S3 The command processing module parses and executes the received commands, generates building information model data or views, a filtering command sequence of viewpoints, and a three-dimensional display command calling sequence, and constructs a building information model data processing and command calling command flow; Executing the data processing and command invoking a command stream, automatically acquiring and querying data related to the command stream, applying the command stream to the relevant data, and sending the command execution result to the three-dimensional display module;

S4 automatically updates the display result of the three-dimensional display module based on the command execution result, so as to realize dynamic interaction with the user and presentation of the display result.
The 3D display method of a voice-controlled building information model according to claim 7, wherein the S2 specifically includes the following steps:

S2.1 uses algorithm modules such as machine learning or deep learning to realize the automatic conversion of user voice input to text, and generate text input corresponding to the voice input;

S2.2 Use statistical models, machine learning models or deep learning models to segment the text;

S2.3 Use natural speech processing models and algorithms to tag the word segmentation results;

S2.4 Text disambiguation: clean up stop words and unify synonyms into a standard word;

In S2.5, based on the part-of-speech tagging in step S2.3 and the text disambiguation result in step S2.4, a natural speech processing algorithm is used to construct a grammar tree of the user input text.

S2.6 Import the data in the three-dimensional display command database and the building information model metadata database, adjust the word segmentation result in step S2.2 according to the data in the database, and compare the word segmentation result and the grammar tree construction result with the read in Perform comparative analysis on the data in the database, check and confirm that each part of the syntax tree can be converted into three-dimensional display control commands and their operation object information; if they cannot be converted into three-dimensional display control commands, terminate the voice control process and wait for the next user Voice input.
The three-dimensional display method of a voice-activated building information model according to claim 7 or 8, wherein the analysis process in step S3 specifically includes the following steps:

S3.1.1 receives the command generated in the voice input module, and imports the data in the three-dimensional display command database and the building information model metadata database;

S3.1.2 Identify and determine which building information model data or views and viewpoints need to be extracted;

S3.1.3 According to the recognition result in step S3.2, generate a data object filtering command sequence corresponding to the user's voice command;

S3.1.4 Identify and determine the three-dimensional display command information that needs to be executed;

S3.1.5 Generate a three-dimensional display command call sequence corresponding to the user's voice command according to the recognition result of the three-dimensional display command in step S3.4;

S3.1.6 Sort the generated data object filtering command sequence and the three-dimensional display command sequence to ensure related commands and execution sequence to form the data processing and command invocation command flow.
The three-dimensional display method of a sound-controlled building information model according to claim 9, wherein the execution process in step S3 specifically includes the following steps:

S3.2.1 Execute the object filtering command: execute the data object filtering command generated in step S3.1.3 in the BIM software, and output the filtered data object list;

S3.2.2 Extract command operation object: group and classify the filtered data objects to form the extracted command operation object;

S3.2.3 Execute 3D display command: execute the corresponding 3D display command for the command operation object;

S3.2.4 Send a 3D display update command to the 3D display module to trigger the effect update of the 3D display module;

S3.2.5 Update the 3D display effect of the device: The 3D display module refreshes the display content according to the 3D display update command sent in step S3.2.4, and completes the execution of the user's voice command.