A digital modular decoration method and system based on building information

Abstract

This invention belongs to the field of digital decoration, and provides a digital modular decoration method and system based on building information technology. The method includes: constructing a digital model of the building to be decorated based on building information modeling technology; automatically generating a modular decoration design scheme that conforms to the building structure and user needs using an autoencoder, wherein the modular decoration design scheme includes functional area division, module size, shape, and material information; determining the required prefabricated decoration modules according to the modular decoration design scheme, generating the installation position and sequence of the prefabricated decoration modules based on the digital model, and uniquely encoding the prefabricated decoration modules based on the installation position and sequence; displaying the digital model and the installation positions of the prefabricated decoration modules on an information terminal, and labeling the corresponding prefabricated decoration module numbers.

Description

Technical Field

[0001] This invention belongs to the field of digital decoration, and specifically relates to a digital modular decoration method and system based on building information. Background Technology

[0002] Modular renovation is a new type of renovation method based on standardization and prefabrication. It breaks down the renovation process into multiple independent functional modules, utilizing factory prefabrication and on-site assembly to achieve efficient, high-quality, and low-cost renovation solutions. Modular renovation has been widely used in residential, commercial, office, and public facilities. Its core objective is to improve renovation efficiency, save costs, reduce resource waste, and meet users' needs for personalization and diverse functions through refined design and manufacturing.

[0003] While modular renovation offers advantages such as efficiency, standardization, and sustainability, the following major technical challenges remain in practical applications: Modular renovation requires design tailored to different building structures and user needs, but current design processes still rely on manual design or rule-based traditional methods; the spatial layout between modules fails to adequately consider spatial topological constraints such as the adjacency of functional areas (e.g., kitchens and dining rooms should be adjacent) and pipeline connectivity (e.g., bathrooms and plumbing systems), leading to logically flawed generated solutions. During the construction phase, insufficient consideration of dimensional, connection, or physical installation constraints between modules can result in practical construction difficulties or mismatches in the generated solutions. Summary of the Invention

[0004] To address the problems in existing technologies, this invention provides a digital modular decoration method based on building information, the method comprising the following steps:

[0005] A digital model of the building to be renovated is constructed based on building information modeling (BIM) technology. The model includes building structure information, spatial layout information, and functional requirements information.

[0006] The self-encoder is used to automatically generate a modular decoration design scheme that meets the building structure and user needs. The modular decoration design scheme includes functional area division, module size, shape and material information.

[0007] Based on the modular decoration design scheme, the required prefabricated decoration modules are determined, and the installation positions and sequences of the prefabricated decoration modules are generated in combination with the digital model. The prefabricated decoration modules are then uniquely coded based on the installation positions and sequences.

[0008] The digital model and the installation location of the prefabricated decoration modules are displayed on the information terminal, and the corresponding prefabricated decoration module numbers are marked.

[0009] Furthermore, the automatic generation of modular interior design schemes that conform to the building structure and user needs using a self-encoder includes:

[0010] Input data includes building structure information, user requirements, and spatial topology.

[0011] The building structure information is extracted from building information model data and represented as a multidimensional tensor, including: geometric features, functional features, and connectivity features;

[0012] The user requirements are represented as a condition vector, including: decoration style, budget constraints, and functional requirements;

[0013] The spatial topology is represented as a graph, including nodes: representing the characteristics of each functional module, and edges: representing the relationships between modules;

[0014] The encoder maps building structure information and user needs to potential spaces:

[0015] Building Information Processing: Geometric features are input into a convolutional network to extract spatial features, and functional features are input into a fully connected layer to generate functional embeddings;

[0016] Spatial relationship modeling: Graph convolutional networks are used to embed the topological graph and output module relationship features;

[0017] User requirement processing: User requirements are input into the embedding layer to generate requirement features;

[0018] Output latent distribution: The encoder outputs the mean and variance of the latent variables, which are used to sample the latent variables;

[0019] The decoder generates interior design schemes from latent variables and user needs:

[0020] Functional area division: Using a conditional embedding layer, a distribution matrix of functional areas is generated based on latent variables;

[0021] Module dimensions: Output the length, width, and height of each module;

[0022] Module Shape: Outputs the geometric shape parameters of the module;

[0023] Material information: The material of the output module.

[0024] Furthermore, the self-encoder includes the following constraints:

[0025] Adjacency constraints ensure the rationality of functional areas between modules;

[0026] Connection constraints ensure that pipelines are correctly connected between modules;

[0027] The spatial distribution of modules should not overlap.

[0028] Furthermore, the step of uniquely numbering the prefabricated decoration modules based on their installation location and sequence includes: extracting the function and location data of each module from the BIM model, connecting the function and location, and automatically generating the unique code by incrementing the number of modules.

[0029] Furthermore, displaying the digital model and the installation location of the prefabricated decoration module on the information terminal, and labeling the corresponding prefabricated decoration module number, includes:

[0030] Exporting a 3D digital model from a BIM system;

[0031] Load the 3D model onto the information terminal;

[0032] Users can click on a module in the model to view detailed module information;

[0033] After selecting a module, the system will highlight the corresponding module in the model and automatically locate the specific installation position;

[0034] The digital model and the installation location of the prefabricated decoration modules are displayed on the information terminal, and the corresponding numbers are marked.

[0035] In another aspect, the present invention provides a digital modular decoration system based on building information, the system comprising the following modules:

[0036] The three-dimensional module is used to construct a digital model of the building to be renovated based on building information modeling technology. The model includes building structure information, spatial layout information, and functional requirements information.

[0037] The generation module is used to automatically generate modular decoration design schemes that conform to the building structure and user needs using an autoencoder. The modular decoration design schemes include functional area division, module size, shape and material information.

[0038] The coding module is used to determine the required prefabricated decoration modules according to the modular decoration design scheme, generate the installation position and sequence of the prefabricated decoration modules in combination with the digital model, and assign a unique code to the prefabricated decoration modules based on the installation position and sequence.

[0039] The display module is used to display the digital model and the installation location of the prefabricated decoration module on the information terminal, and to label the corresponding prefabricated decoration module number.

[0040] Furthermore, the automatic generation of modular interior design schemes that conform to the building structure and user needs using a self-encoder includes:

[0041] Input data includes building structure information, user requirements, and spatial topology.

[0042] The building structure information is extracted from building information model data and represented as a multidimensional tensor, including: geometric features, functional features, and connectivity features;

[0043] The user requirements are represented as a condition vector, including: decoration style, budget constraints, and functional requirements;

[0044] The spatial topology is represented as a graph, including nodes: representing the characteristics of each functional module, and edges: representing the relationships between modules;

[0045] The encoder maps building structure information and user needs to potential spaces:

[0046] Building Information Processing: Geometric features are input into a convolutional network to extract spatial features, and functional features are input into a fully connected layer to generate functional embeddings;

[0047] Spatial relationship modeling: Graph convolutional networks are used to embed the topological graph and output module relationship features;

[0048] User requirement processing: User requirements are input into the embedding layer to generate requirement features;

[0049] Output latent distribution: The encoder outputs the mean and variance of the latent variables, which are used to sample the latent variables;

[0050] The decoder generates interior design schemes from latent variables and user needs:

[0051] Functional area division: Using a conditional embedding layer, a distribution matrix of functional areas is generated based on latent variables;

[0052] Module dimensions: Output the length, width, and height of each module;

[0053] Module Shape: Outputs the geometric shape parameters of the module;

[0054] Material information: The material of the output module.

[0055] Furthermore, the self-encoder includes the following constraints:

[0056] Adjacency constraints ensure the rationality of functional areas between modules;

[0057] Connection constraints ensure that pipelines are correctly connected between modules;

[0058] The spatial distribution of modules should not overlap.

[0059] Furthermore, the step of uniquely numbering the prefabricated decoration modules based on their installation location and sequence includes: extracting the function and location data of each module from the BIM model, connecting the function and location, and automatically generating the unique code by incrementing the number of modules.

[0060] Furthermore, displaying the digital model and the installation location of the prefabricated decoration module on the information terminal, and labeling the corresponding prefabricated decoration module number, includes:

[0061] Exporting a 3D digital model from a BIM system;

[0062] Load the 3D model onto the information terminal;

[0063] Users can click on a module in the model to view detailed module information;

[0064] After selecting a module, the system will highlight the corresponding module in the model and automatically locate the specific installation position;

[0065] The digital model and the installation location of the prefabricated decoration modules are displayed on the information terminal, and the corresponding numbers are marked.

[0066] The modular decoration design method of this invention automatically generates decoration design schemes that conform to the building structure and user needs through an autoencoder, significantly improving design efficiency and scheme quality. It also solves core technical problems in traditional modular decoration, and has the following specific beneficial effects:

[0067] Based on a conditional generation method, it supports real-time adjustments to user requirements (such as style, budget, and functional zoning) to quickly generate high-quality customized solutions. Graph neural networks are used to model the adjacency and connectivity between modules, ensuring the logical consistency of the generated solutions in spatial layout and functional zoning. During solution generation, it automatically detects and optimizes dimensional conflicts, pipeline connections, and spatial overlaps between modules, ensuring the solutions meet the feasibility and consistency of actual construction. Multi-objective optimization constraints ensure that functional zoning (such as kitchen and dining room, bedroom and bathroom) is rationally laid out and meets actual usage needs. Pipeline connection paths (such as water, electricity, and gas) between modules are clearly defined during the design phase, avoiding additional adjustments due to design flaws during later construction. It automatically detects and eliminates dimensional and positional conflicts between modules, ensuring that the generated solutions can be directly used for construction, improving actual execution efficiency.

[0068] This invention effectively solves the technical problems of low design efficiency, unstable scheme quality, lack of personalized and diverse support, and poor construction feasibility in current modular decoration through innovative generative models and optimization methods, and provides an intelligent, efficient and high-quality technical solution for modular decoration. Attached Figure Description

[0069] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0070] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0071] The invention will now be described in preferred form with reference to the accompanying drawings and specific embodiments.

[0072] This embodiment solves the above problems through the following steps:

[0073] In one embodiment, reference Figure 1 This invention provides a digital modular decoration method based on building information technology. By combining Building Information Modeling (BIM) technology with modular decoration design technology, it digitally constructs an information model of the building decoration process. Based on this information model, the decoration scheme is modularly decomposed and optimized, thereby achieving efficient collaboration and full-process digital management of decoration design, module production, on-site construction, and subsequent operation and maintenance. Specifically, it includes the following steps:

[0074] Step S10: Construct a digital model of the building to be renovated based on Building Information Modeling (BIM) technology. The model includes building structure information, spatial layout information, and functional requirements information.

[0075] In this step, physical structural data of the building to be renovated is obtained through laser scanning, drone mapping or existing design drawings, including the geometric shape and size information of the building's walls, floors, columns and beams, and information on the layout of the building's functional areas, ventilation ducts, electrical wiring and water supply and drainage pipes is collected.

[0076] Building Information Modeling (BIM) technology is used to integrate the collected physical information of the building with the decoration requirements to create a three-dimensional digital model. The model includes not only the geometric shape of the building structure, but also its attribute information, such as component material type, strength, durability parameters, etc.

[0077] Based on the user's decoration needs, functional area division information is defined and embedded. The functional needs include the use of the building space (such as living room, bedroom, kitchen, etc.), functional layout requirements (such as furniture arrangement, equipment installation location), and decoration style preferences (such as modern, minimalist, retro, etc.).

[0078] The initially established digital model is checked and refined to ensure the completeness and accuracy of the model's geometric precision, functional requirements description, and attribute information;

[0079] The completed digital model is stored as a format file that supports multi-dimensional information. This file can support the parsing and automatic generation of subsequent modular decoration designs, ensuring that all steps in the decoration process are based on the digital model for collaborative operation.

[0080] Step S20: Use an autoencoder to automatically generate a modular decoration design scheme that conforms to the building structure and user needs. The modular decoration design scheme includes functional area division, module size, shape and material information.

[0081] In this step, building information model (BIM) data and user requirements are combined to generate a modular interior design scheme that includes functional area division, module dimensions, module shapes, and material information. The specific process is as follows:

[0082] Input data includes building structure information, user requirements, and spatial topology.

[0083] Building structural information is extracted from BIM data and represented as a multidimensional tensor, including the following:

[0084] Geometric features: Three-dimensional spatial features (such as the geometry and dimensions of walls, floors, columns, and beams).

[0085] Functional characteristics: Initial zoning information for functional areas, such as living room, kitchen, bedroom, etc.

[0086] Connection characteristics: Information on the connection of internal building pipelines, such as electrical circuits, water pipes, ventilation systems, etc.

[0087] User requirements are represented as a condition vector, including:

[0088] Interior design style: such as modern, minimalist, retro.

[0089] Budget limit: The total budget acceptable to the user.

[0090] Functional requirements: Functional requirements for module division (e.g., the bedroom must be close to the bathroom).

[0091] The spatial relationships between modules are represented by the following diagram:

[0092] Nodes: Represent the characteristics of each functional module, such as the location of the center point of each module.

[0093] Edge: Represents the relationship between modules, such as the adjacency between module i and module j.

[0094] The model structure includes an encoder and a decoder.

[0095] The encoder maps building structure information and user needs to potential spaces:

[0096] Building Information Processing:

[0097] Geometric features are input into a convolutional network to extract spatial features.

[0098] Functional features are input into the fully connected layer to generate functional embeddings.

[0099] Spatial Relationship Modeling:

[0100] Graph Convolutional Networks (GCNs) are used to embed the topological graph and output module relationship features.

[0101] User request handling:

[0102] User requirements are input into the embedding layer to generate requirement features.

[0103] Output latent distribution:

[0104] The encoder outputs the mean and variance of the latent variables, which are used to sample the latent variables.

[0105] The decoder generates interior design schemes from latent variables and user needs:

[0106] Functional area division:

[0107] Using a conditional embedding layer, the distribution matrix of functional areas is generated based on latent variables.

[0108] Such as the center point of the living room, kitchen, and bedroom.

[0109] Module dimensions:

[0110] Output the length, width, and height of each module.

[0111] For example, the size of the bedroom module.

[0112] Module shape:

[0113] Output the geometric parameters of the module, such as the thickness of the wall and the size of the window.

[0114] Material information:

[0115] The material of the output module, such as the surface material of a wall.

[0116] To obtain the correct solution, the following constraints also need to be applied during the coding process:

[0117] Adjacency constraints ensure the rationality of functional areas between modules:

[0118]

[0119] Among them, L adjacency This indicates the adjacency constraint value, which is not greater than a preset threshold; d ij d represents the distance between module i and module j, and E represents the set of modules; max This indicates the maximum allowed distance between adjacent modules.

[0120] Connection constraints ensure that the pipelines between modules are correctly connected:

[0121]

[0122] Among them, L connectivity C represents the connection constraint value. i C represents the connection point of pipeline i. j Let ||||2 represent the connection point of pipeline j, and ||||2 represent the L2 normal form.

[0123] The spatial distribution between modules should not overlap:

[0124]

[0125] Among them, L overlap The overlapArea(i,j) represents the amount of spatial overlap, where OverlapArea(i,j) represents the overlap area between modules i and j.

[0126] The generated modular interior design scheme includes the following information:

[0127] Functional zone division: Location, coordinates of the center point of the functional zone; Distribution, coordinates of the boundaries of the functional zone.

[0128] Module dimensions: length, width, and height; physical dimensions of each module; module shape:

[0129] Geometric parameters: such as the curvature of a curved wall.

[0130] Material information: such as the materials used for the walls.

[0131] In this step, the encoder processes the input data to learn the distribution characteristics of architectural information and user needs, and establishes connections within the potential space. This learning mechanism ensures that the functional division and spatial layout of modules are based on the building structure; the generated potential space captures a variety of potential design possibilities.

[0132] The decoder combines user requirements (such as style and budget) to generate conditional outputs in the latent space, ensuring that the output scheme meets the constraints specified by the user.

[0133] Graph neural networks model the adjacency and connectivity characteristics of modules, ensuring that the generated schemes meet the physical constraints between functional areas and modules.

[0134] By modeling the geometric features, functional area features, and spatial relationships of BIM data, key information in the building structure was captured.

[0135] Using user demand vectors as conditional variables as input ensures the personalization and controllability of the generated solutions.

[0136] The sampling mechanism of the latent space can generate diverse schemes, while the decoder ensures that these schemes meet the actual constraints.

[0137] Step S30: Based on the modular decoration design scheme, determine the required prefabricated decoration modules, generate the installation positions and sequences of the prefabricated decoration modules in conjunction with the digital model, and assign a unique number to the prefabricated decoration modules based on the installation positions and sequences.

[0138] Extracting module information from modular interior design schemes

[0139] For example, the modular interior design scheme shows that the living room area requires the following modules:

[0140] Wall modules:

[0141] Dimensions: 2.5 meters long, 0.2 meters wide, and 3 meters high;

[0142] Materials: Concrete + white paint;

[0143] Function: Partition wall with concealed wiring channels.

[0144] Floor modules:

[0145] Dimensions: 0.6 meters × 0.6 meters;

[0146] Material: Ceramic tile (gray, non-slip).

[0147] Ceiling module:

[0148] Size: 2 meters x 2 meters per piece;

[0149] Materials: Lightweight aluminum alloy, integrated LED lights.

[0150] Modules are categorized by function and region:

[0151] Wall modules:

[0152] Two standard partition walls are used to separate the living room and dining room;

[0153] A wall panel with cable routing channels, used as a TV background wall.

[0154] Floor modules:

[0155] The living room needs 30 non-slip ceramic tile flooring modules (total area 18 square meters).

[0156] Ceiling module:

[0157] Four integrated lighting modules are needed (covering 8 square meters).

[0158] Check that each module meets the prefabrication requirements:

[0159] The wall modules are divided into standard-sized units, each 2.5 meters long and 3 meters high;

[0160] The floor and ceiling modules require no adjustment and can be directly adopted in standard size.

[0161] Obtain the specific installation location and area of ​​the module from the BIM model:

[0162] Wall modules:

[0163] The partition wall module is located in the straight line area from the living room entrance to the left side of the dining room;

[0164] The start and end points of installation for each wall module are directly marked in the BIM model, for example:

[0165] Block 1: Start point (0,0,0), End point (2.5,0,3).

[0166] Floor modules:

[0167] The floor modules start from the left side of the living room and are arranged in a grid pattern;

[0168] The installation sequence is from left to right and from front to back, and the coordinates of each floor module are clearly marked in the BIM model.

[0169] Ceiling module:

[0170] The ceiling module with integrated LED lights is installed in the center of the living room to ensure even distribution.

[0171] Based on the module's regional and functional divisions, determine a reasonable installation sequence:

[0172] Wall modules preferred:

[0173] The walls need to be installed first to ensure that the space is divided before proceeding with subsequent operations.

[0174] The installation order is divided by area; first install the partition wall, then install the feature wall.

[0175] Floor modules are the second option:

[0176] Lay the flooring in grid order to avoid damage during subsequent construction.

[0177] Ceiling module final:

[0178] The ceiling and lighting fixtures will be installed after the wall and floor modules are installed.

[0179] Modules are uniquely numbered using a combination of module function, region, and sequence:

[0180] Numbering format: [Function Code]-[Region Code]-[Sequence Number].

[0181] Example:

[0182] Wall module number: WT-01-001

[0183] Function code: WT (Wall Module);

[0184] Area code: 01 (living room);

[0185] Serial number: 001 (partition wall, first piece).

[0186] Floor module number: FL-01-015

[0187] Function code: FL (Floor module);

[0188] Area code: 01 (living room);

[0189] Serial number: 015 (15th tile).

[0190] Ceiling module number: CL-01-003

[0191] Function code: CL (ceiling module);

[0192] Area code: 01 (living room);

[0193] Serial number: 003 (3rd ceiling panel).

[0194] The system extracts the function and location data of each module from the BIM model and automatically generates a number.

[0195] Example of generated number:

[0196] Wall modules:

[0197] Piece 1: WT-01-001

[0198] Part 2: WT-01-002

[0199] Floor modules:

[0200] Part 1: FL-01-001

[0201] Block 30: FL-01-030

[0202] Ceiling module:

[0203] Part 1: CL-01-001

[0204] Part 4: CL-01-004

[0205] The factory produces the corresponding modules according to the serial numbers and marks the serial numbers on the surface of the modules to facilitate transportation and installation.

[0206] For example, after the wall module WT-01-001 is manufactured, this number will be marked, along with the installation area and location.

[0207] Modules are packaged and transported according to their serial numbers, and arranged in the order of installation during warehousing to reduce on-site search time.

[0208] Workers install the equipment according to the numbered list and the installation locations in the BIM model to ensure rapid matching:

[0209] For example, if you look at the number WT-01-001, the corresponding location in the BIM model is the wall installation at the entrance of the living room.

[0210] Step S40: Display the digital model and the installation location of the prefabricated decoration module on the information terminal, and label the corresponding prefabricated decoration module number.

[0211] The digital model and installation locations of prefabricated decoration modules are displayed on the information terminal, with the corresponding prefabricated decoration module numbers clearly marked, ensuring that construction personnel can accurately locate and quickly match modules during installation. The specific implementation method and detailed description are as follows:

[0212] Export a 3D digital model from the BIM system, including building structure information, module division, functional areas, etc.

[0213] Load 3D models onto information terminals (such as tablets, smartphones, or AR / VR devices);

[0214] It offers multi-view functionality, including top view, cross-sectional view, and local zoom.

[0215] For example, the information terminal displays a complete building model, in which the living room area is marked with the division of wall, floor and ceiling modules.

[0216] Users can click on a module in the model to view detailed module information, including dimensions, materials, functions, and serial numbers.

[0217] After selecting a module, the system will highlight the corresponding module in the model and automatically locate the specific installation position.

[0218] When you click on the wall module, the wall's dimensions (2.5m × 3m), material (concrete), and number (WT-01-001) will pop up on the side of the screen.

[0219] The digital model and the installation location of the prefabricated decoration modules are displayed on the information terminal, and the corresponding numbers are marked. Construction workers can quickly locate the modules and understand the installation steps, which significantly improves construction efficiency and reduces installation errors.

[0220] On the other hand, the present invention also provides a digital modular decoration system based on building information, comprising:

[0221] The three-dimensional module is used to construct a digital model of the building to be renovated based on building information modeling technology. The model includes building structure information, spatial layout information, and functional requirements information.

[0222] The generation module is used to automatically generate modular decoration design schemes that conform to the building structure and user needs using an autoencoder. The modular decoration design schemes include functional area division, module size, shape and material information.

[0223] The coding module is used to determine the required prefabricated decoration modules according to the modular decoration design scheme, generate the installation position and sequence of the prefabricated decoration modules in combination with the digital model, and assign a unique code to the prefabricated decoration modules based on the installation position and sequence.

[0224] The display module is used to display the digital model and the installation location of the prefabricated decoration module on the information terminal, and to label the corresponding prefabricated decoration module number.

[0225] For any module structures not specifically defined in this invention, the existing technical descriptions shall prevail. The prior art mentioned in the foregoing background and specific embodiments sections can be considered part of this invention and used to understand the meaning of certain technical features or parameters.

Claims

1. A digital modular decoration method based on building information, characterized in that, The method includes the following steps: A digital model of the building to be renovated is constructed based on building information modeling (BIM) technology. The model includes building structure information, spatial layout information, and functional requirements information. The self-encoder is used to automatically generate a modular decoration design scheme that meets the building structure and user needs. The modular decoration design scheme includes functional area division, module size, shape and material information. Based on the modular decoration design scheme, the required prefabricated decoration modules are determined, and the installation positions and sequences of the prefabricated decoration modules are generated in combination with the digital model. The prefabricated decoration modules are then uniquely coded based on the installation positions and sequences. The information terminal displays the digital model and the installation location of the prefabricated decoration modules, and labels the corresponding prefabricated decoration module numbers; the automatic generation of modular decoration design schemes that conform to the building structure and user needs using an autoencoder includes: Input data includes building structure information, user requirements, and spatial topology. The building structure information is extracted from building information model data and represented as a multidimensional tensor, including: geometric features, functional features, and connectivity features; The user requirements are represented as a condition vector, including: decoration style, budget constraints, and functional requirements; The spatial topology is represented as a graph, including nodes: representing the characteristics of each functional module, and edges: representing the relationships between modules; The encoder maps building structure information and user needs to a latent space: Building Information Processing: Geometric features are input into a convolutional network to extract spatial features, and functional features are input into a fully connected layer to generate functional embeddings; Spatial relationship modeling: Graph convolutional networks are used to embed the topological graph and output module relationship features; User requirement processing: User requirements are input into the embedding layer to generate requirement features; Output latent distribution: The encoder outputs the mean and variance of the latent variables, which are used to sample the latent variables; The decoder generates interior design schemes from latent variables and user needs: Functional area division: Using a conditional embedding layer, a distribution matrix of functional areas is generated based on latent variables; Module dimensions: Output the length, width, and height of each module; Module Shape: Outputs the geometric shape parameters of the module; Material information: The material of the output module; The following constraints also need to be applied during the coding process: Adjacency constraints ensure the rationality of functional areas between modules: Among them, L adjacency This indicates the adjacency constraint value, which is not greater than a preset threshold; d ij d represents the distance between module i and module j, and E represents the set of modules; max Indicates the maximum allowed distance between adjacent modules; Connection constraints ensure that the pipelines between modules are correctly connected: Among them, L connectivity C represents the connection constraint value. i C represents the connection point of pipeline i. j This represents the connection point of pipeline j, and || ||2 represents the L2 normal form; The spatial distribution between modules should not overlap: Among them, L overlap The overlapArea(i,j) represents the amount of spatial overlap, where OverlapArea(i,j) represents the overlap area between modules i and j.

2. The digital modular decoration method based on building information according to claim 1, characterized in that, The autoencoder includes the following constraints: Adjacency constraints ensure the rationality of functional areas between modules; Connection constraints ensure that pipelines are correctly connected between modules; The spatial distribution of modules should not overlap.

3. The digital modular decoration method based on building information according to claim 1, characterized in that, The step of uniquely numbering the prefabricated decoration modules based on their installation location and sequence includes: extracting the function and location data of each module from the BIM model, connecting the function and location, and automatically generating the unique code by incrementing the number of modules.

4. The digital modular decoration method based on building information according to claim 1, characterized in that, The step of displaying the digital model and the installation location of the prefabricated decoration modules on the information terminal, and labeling the corresponding prefabricated decoration module numbers, includes: Exporting a 3D digital model from a BIM system; Load the 3D model onto the information terminal; Users can click on a module in the model to view detailed module information; After selecting a module, the system will highlight the corresponding module in the model and automatically locate the specific installation position; The digital model and the installation location of the prefabricated decoration modules are displayed on the information terminal, and the corresponding numbers are marked.

5. A digital modular decoration system based on building information, characterized in that, The system includes the following modules: The three-dimensional module is used to construct a digital model of the building to be renovated based on building information modeling technology. The model includes building structure information, spatial layout information, and functional requirements information. The generation module is used to automatically generate modular decoration design schemes that conform to the building structure and user needs using an autoencoder. The modular decoration design schemes include functional area division, module size, shape and material information. The coding module is used to determine the required prefabricated decoration modules according to the modular decoration design scheme, generate the installation position and sequence of the prefabricated decoration modules in combination with the digital model, and assign a unique code to the prefabricated decoration modules based on the installation position and sequence. The display module is used to display the digital model and the installation location of the prefabricated decoration module on the information terminal, and to label the corresponding prefabricated decoration module number. The method of automatically generating modular interior design schemes that conform to building structure and user needs using an autoencoder includes: Input data includes building structure information, user requirements, and spatial topology. The building structure information is extracted from building information model data and represented as a multidimensional tensor, including: geometric features, functional features, and connectivity features; The user requirements are represented as a condition vector, including: decoration style, budget constraints, and functional requirements; The spatial topology is represented as a graph, including nodes: representing the characteristics of each functional module, and edges: representing the relationships between modules; The encoder maps building structure information and user needs to potential spaces: Building Information Processing: Geometric features are input into a convolutional network to extract spatial features, and functional features are input into a fully connected layer to generate functional embeddings; Spatial relationship modeling: Graph convolutional networks are used to embed the topological graph and output module relationship features; User requirement processing: User requirements are input into the embedding layer to generate requirement features; Output latent distribution: The encoder outputs the mean and variance of the latent variables, which are used to sample the latent variables; The decoder generates interior design schemes from latent variables and user needs: Functional area division: Using a conditional embedding layer, a distribution matrix of functional areas is generated based on latent variables; Module dimensions: Output the length, width, and height of each module; Module Shape: Outputs the geometric shape parameters of the module; Material information: The material of the output module; The following constraints also need to be applied during the coding process: Adjacency constraints ensure the rationality of functional areas between modules: Among them, L adjacency This indicates the adjacency constraint value, which is not greater than a preset threshold; d ij d represents the distance between module i and module j, and E represents the set of modules; max Indicates the maximum allowed distance between adjacent modules; Connection constraints ensure that the pipelines between modules are correctly connected: Among them, L connectivity C represents the connection constraint value. i C represents the connection point of pipeline i. j This represents the connection point of pipeline j, and || ||2 represents the L2 normal form; The spatial distribution between modules should not overlap: Among them, L overlap The overlapArea(i,j) represents the amount of spatial overlap, where OverlapArea(i,j) represents the overlap area between modules i and j.

6. The digital modular decoration system based on building information according to claim 5, characterized in that, The autoencoder includes the following constraints: Adjacency constraints ensure the rationality of functional areas between modules; Connection constraints ensure that pipelines are correctly connected between modules; The spatial distribution of modules should not overlap.

7. The digital modular decoration system based on building information according to claim 5, characterized in that, The step of uniquely numbering the prefabricated decoration modules based on their installation location and sequence includes: extracting the function and location data of each module from the BIM model, connecting the function and location, and automatically generating the unique code by incrementing the number of modules.

8. The digital modular decoration system based on building information according to claim 5, characterized in that, The step of displaying the digital model and the installation location of the prefabricated decoration modules on the information terminal, and labeling the corresponding prefabricated decoration module numbers, includes: Exporting a 3D digital model from a BIM system; Load the 3D model onto the information terminal; Users can click on a module in the model to view detailed module information; After selecting a module, the system will highlight the corresponding module in the model and automatically locate the specific installation position; The digital model and the installation location of the prefabricated decoration modules are displayed on the information terminal, and the corresponding numbers are marked.

Images