A three-dimensional digital survey and design method and system based on BIM

Abstract

The present invention discloses a BIM-based three-dimensional digital survey and design method and system, relating to the technical field of building survey and design. A BIM-based three-dimensional digital survey and design system includes: a data acquisition module, a component recognition module, a physical performance analysis module, a BIM modeling module, a reverse adjustment module, a performance analysis and optimization module, and a user interaction module. The present invention automatically identifies and classifies the main components in a building through a component recognition model, extracts the geometric and physical properties of the components, and establishes a component property database. The model has the ability to classify skeletal components and functional components, and retains the marker points required for alignment, achieving accurate mapping from point clouds to structural semantics, significantly reducing the burden of manual modeling.

Description

Technical Field

[0001] The present invention relates to the technical field of building survey and design, and in particular to a BIM-based three-dimensional digital survey and design method and system. Background Art

[0002] As the construction industry continues its digital and intelligent transformation, Building Information Modeling (BIM) technology is increasingly being used in building design, construction, and operations. BIM integrates a building's three-dimensional spatial information, component properties, construction processes, and management requirements into a complete information model, providing a comprehensive, visual digital platform for building design, construction, and operations. However, in practice, design methods based on traditional manual surveying and two-dimensional drawings still face challenges, particularly in terms of architectural design accuracy and efficiency.

[0003] Traditional architectural survey methods rely primarily on manual measurement, 2D drawings, and hand-modeling. These methods are not only susceptible to human error but also struggle to quickly and accurately obtain true 3D building data. Furthermore, existing architectural survey and design solutions often lack sufficient digitalization and intelligence, making it difficult to effectively evaluate and optimize the various complex factors in the building design process, such as structural safety, energy consumption, and daylighting.

[0004] Building on this foundation, three-dimensional digital design methods based on BIM technology have emerged in recent years. By combining building point cloud data with BIM models, full lifecycle management, from surveying and design to construction, can be achieved. However, existing BIM-based design methods still have several limitations, primarily low automation, insufficient component recognition accuracy, a lack of structural analysis for model optimization, and a lack of dynamic adjustment and optimization mechanisms. Summary of the Invention

[0005] The present invention proposes a three-dimensional digital survey and design method and system based on BIM, which automatically acquires components and generates a skeleton structure model by combining a three-dimensional spatial data model. The BIM model is automatically generated and rationally optimized through the skeleton structure and inverse algorithm of the BIM model to solve the above problems.

[0006] A BIM-based three-dimensional digital survey and design method, comprising:

[0007] Obtain the point cloud data and corresponding material information of the target building, and perform cleaning, coordinate registration, and splicing to generate an initial 3D spatial data model;

[0008] Based on the three-dimensional spatial data model, a component recognition model is used to automatically identify and classify the main components in the building, and the geometric and physical properties of the components are extracted to establish a component attribute database; the classification includes skeleton components and functional components; the components retain two marking points for alignment;

[0009] Performing physical property analysis on the bone components in combination with the corresponding material information, and splicing the bone components into a bone structure model based on the physical property analysis results;

[0010] The skeletal structure model and functional components are aligned and data replaced in the three-dimensional spatial data model; the preset BIM component family template is called to automatically construct the three-dimensional spatial data model after data replacement into a BIM model, and the architecture corresponding to the skeletal structure model in the BIM model is marked as the skeletal architecture of the BIM model; the process of automatically constructing the BIM model complies with architectural design standards and specifications;

[0011] Combine functional components to obtain design requirements, make component-level adjustments to the existing BIM model based on preset design rules and reverse deduction algorithms, and generate the physical performance requirements of the BIM model. After the adjustments, perform physical performance analysis on the skeleton structure to ensure that it meets the physical performance requirements of the BIM model.

[0012] As a preferred technical solution of the present invention, a BIM-based three-dimensional digital survey and design method further includes:

[0013] Perform a comprehensive performance analysis on the adjusted BIM model, including structural mechanical performance, energy consumption simulation, and natural lighting. Match optimization recommendations with corresponding design requirements based on the comprehensive performance analysis results to drive automatic iterative updates of the BIM model.

[0014] It supports users to dynamically adjust design requirements, and displays various BIM models and their comprehensive performance analysis results obtained during the optimization process through a visual interface, enabling rapid "simulation hypothesis" analysis and scheme comparison, helping users make decisions among multiple design schemes.

[0015] As a preferred technical solution of the present invention, the structure of the component identification model includes:

[0016] Input layer, used to obtain the three-dimensional spatial data model of the building;

[0017] Feature extraction layer, used to extract multi-scale spatial features of point cloud data in the three-dimensional spatial data model and construct a feature pyramid;

[0018] The component segmentation layer is used to segment the entire building according to the feature pyramid, generate multiple unclassified components, and mark each component with two spatial marker points for subsequent alignment;

[0019] The attribute analysis layer is used to extract the geometric properties of each component and obtain the physical properties of the component in combination with the corresponding material information;

[0020] The component classification layer is used to perform functional analysis and classification of unclassified components. It adopts a staged classification method, first identifying all bone components and then identifying functional components;

[0021] The adjacency association analysis layer is used to analyze the spatial adjacency and structural connection relationships between components, and to correct the results of the component classification layer based on the association information;

[0022] The classification output layer is used to output the final identified component category, geometric properties and physical properties to the component property database.

[0023] As a preferred technical solution of the present invention, the training of the component recognition model includes:

[0024] A three-dimensional spatial data model of a building with complete annotated skeletal and functional components is obtained as a data sample, and an initial training set is constructed. The three-dimensional convolutional neural network is pre-trained using a few-shot learning method combined with the initial training set to obtain an initial component recognition model. A self-training mechanism is used to perform incremental training in combination with a large number of unannotated three-dimensional spatial data models of buildings. The parameters of the initial component recognition model are optimized through pseudo-label generation and confidence screening mechanisms to improve the recognition accuracy of different types of components. Component physical properties and material information are introduced as auxiliary features during the training process to optimize the discrimination boundary of component classification and improve the distinction between functional and skeletal components. The training results are evaluated for accuracy through cross-validation and structural adjacent relationship analysis, and the optimal model parameter set for component recognition is finally determined to obtain the final component recognition model.

[0025] As a preferred technical solution of the present invention, the bone structure model includes:

[0026] The geometric properties and material information of all bone components are obtained, and the physical properties of each bone component are analyzed based on the finite element analysis method. The physical properties analysis includes load-bearing capacity, flexural stiffness, node connection performance and material strength indicators. Based on the analysis results, bone component pairs that meet the requirements of structural safety and connection feasibility are selected, and a bone component connection map is constructed based on the spatial position relationship and mechanical matching relationship of the components. The edge weights in the connection map reflect the connection strength and component coupling degree. Combining the component connection map with the building structure design specifications, the bone components are topologically spliced ​​using a graph traversal algorithm and mechanical consistency constraints to generate an initial bone structure model. Based on the initial bone structure model, the overall structure finite element analysis is further performed to evaluate the overall stability and resistance performance of the spliced ​​structure, and the component connection relationships that do not meet the performance requirements are locally optimized and adjusted until a bone structure model that meets the target physical performance requirements is generated.

[0027] As a preferred technical solution of the present invention, the construction of the BIM model includes: based on the replaced three-dimensional space data model, automatically matching and replacing the components in the model by calling the preset BIM component family template, the component family template includes the geometry, physical properties and construction process of the building components; automatically adjusting the component size, position and connection method according to the architectural design standards and specifications, and generating a complete BIM model, and marking the corresponding content of the bone structure model in the BIM model as the bone structure.

[0028] As a preferred technical solution of the present invention, the component-level adjustment includes:

[0029] The system obtains the design goals input by the user, calls the goal-driven rules and component constraint rules in the design rule library, establishes an optimization function based on the multi-objective optimization algorithm, and generates component-level change suggestions that meet the design goals through reverse calculation. The component adjustment process is combined with the performance impact assessment mechanism to evaluate the impact of component changes on structural performance, energy consumption performance or lighting efficiency in real time, and dynamically adjust the optimization strategy accordingly.

[0030] A BIM-based three-dimensional digital survey and design system, including:

[0031] The data acquisition module obtains the point cloud data and corresponding material information of the target building and generates an initial three-dimensional spatial data model;

[0032] The component recognition module uses a component recognition model based on a three-dimensional spatial data model to automatically identify and classify the main components in the building, extract the geometric and physical properties of the components, and establish a component property database;

[0033] The physical performance analysis module analyzes the physical properties of bone components based on material information and assembles the bone components into a bone structure model based on the analysis results;

[0034] The BIM modeling module aligns and replaces the bone structure model and functional components in the three-dimensional spatial data model, calls the preset BIM component family template, automatically generates the BIM model, and obtains the bone structure;

[0035] The reverse adjustment module adjusts the existing BIM model at the component level based on preset design rules and reverse inference algorithms, generates the physical performance requirements of the BIM model, and analyzes the skeleton to ensure that it meets the physical performance requirements of the BIM model;

[0036] The performance analysis and optimization module conducts comprehensive performance analysis on the adjusted BIM model, matches optimization suggestions based on the analysis results, and drives automatic iterative updates of the BIM model;

[0037] The user interaction module supports users to dynamically adjust design requirements and displays various BIM models and their comprehensive performance analysis results during the optimization process through a visual interface.

[0038] The present invention has the following advantages:

[0039] The present invention uses a component recognition model to automatically identify and classify the main components in a building, extract the geometric and physical properties of the components, and establish a component attribute database. The model has the ability to classify skeletal components and functional components, and retains the marking points required for alignment, achieving accurate mapping from point clouds to structural semantics, significantly reducing the burden of manual modeling.

[0040] The present invention uses physical performance analysis to splice bone components to form a bone structure model, which serves as the "bone structure" for subsequent BIM model construction. This method implements a BIM modeling process driven by structural performance, ensures the structural safety and performance controllability of the building model, and realizes "form-structure-performance" integrated modeling.

[0041] Based on the alignment of the bone structure model and functional components, the present invention calls the preset BIM component family template to realize the automatic replacement of components and standardized reconstruction of the model structure, automatically generates a BIM model that complies with building specifications, and significantly improves modeling efficiency and standard consistency.

[0042] This invention supports component-level adjustments to existing BIM models based on pre-set design rules and inverse algorithms, combining functional components with user design requirements. This process not only considers design objectives but also optimizes and analyzes physical performance requirements, enabling intelligent feedback adjustments from performance to form. The adjusted BIM model undergoes a comprehensive performance analysis, generating optimization recommendations based on the results. This drives automatic iterative updates of the BIM model, significantly improving the scientific and rational nature of architectural design. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only schematic diagrams of the present invention. Those skilled in the art can also derive other drawings based on the provided drawings without inventive effort.

[0044] Figure 1 This is a structural diagram of a BIM-based three-dimensional digital survey and design system adopted in an embodiment of the present invention. DETAILED DESCRIPTION

[0045] To make the objectives, technical solutions, and advantages of the present invention more clear, the present invention will be further described in detail below with reference to the accompanying drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0046] Example 1, a BIM-based three-dimensional digital survey and design method, comprising the following steps:

[0047] Step S1: Obtain the point cloud data and corresponding material information of the target building, and perform cleaning, coordinate registration, and splicing processing to generate an initial three-dimensional spatial data model;

[0048] S101 point cloud data acquisition:

[0049] Using laser scanning equipment (SLAM systems) or drone-mounted 3D scanners, a panoramic scan of the target building is performed to obtain raw point cloud data covering the entire structure. During the scanning process, at least the spatial coordinates, reflection intensity, and RGB color value of each acquisition point are recorded and numbered and categorized according to the scanning site. Combining manual sampling records, architectural drawing analysis, and IoT sensors, the material type, material number, and mechanical performance parameters (density, elastic modulus, thermal conductivity) of each area of ​​the building are collected and annotated. A material property information table is constructed and matched with the point cloud data by spatial region.

[0050] S102, pre-processing the original point cloud data, including:

[0051] Noise removal (isolated point filtering, voxel filtering);

[0052] Data compression and resampling (downsampling based on voxel grid);

[0053] Identify and eliminate invalid and duplicate points;

[0054] Remove non-building related background data (ground debris, moving objects).

[0055] S103, coordinate registration processing:

[0056] In the case of multi-site scanning or multi-view acquisition, a marker-based geometric registration algorithm is used to spatially align the point clouds of each sub-region; the registered point cloud data of each region are spatially spliced ​​and fused to generate a complete building point cloud model; large-scale point cloud data are divided into multiple data blocks according to building areas or functional areas to improve subsequent processing efficiency; point cloud data is structured and stored to form an initial three-dimensional spatial data model that can be used for subsequent component identification.

[0057] Output the structured format (.pcd) of the 3D spatial data model; spatially map and associate point cloud data with material property data to provide basic data support for subsequent component identification and physical performance analysis.

[0058] Step S2: Based on the three-dimensional spatial data model, a component recognition model is used to automatically identify and classify the main components in the building, and the geometric and physical properties of the components are extracted to establish a component attribute database; the classification includes skeleton components and functional components; the components retain two marking points for alignment;

[0059] S201, component identification model construction and input preprocessing:

[0060] Based on the three-dimensional spatial data model output from step S1, it is input into the preset component recognition model; the point cloud data is normalized, and a layered sampling strategy is adopted to construct data views of different density levels; the point cloud is divided into processable sub-areas according to spatial blocks, the input dimension of the neural network model is adapted, and a multi-scale feature input structure is constructed.

[0061] S202, the structure of the component identification model:

[0062] The component recognition model is a multi-module recognition system based on a fusion architecture of three-dimensional convolutional neural networks (3D-CNN) and graph neural networks (GNN), specifically including:

[0063] Input layer: receives the normalized three-dimensional spatial data model;

[0064] Feature extraction layer: extracts local and global spatial features of the point cloud through voxelization and PointNet++ network, and constructs a feature pyramid;

[0065] Component segmentation layer: Combining a spatial clustering algorithm (DBSCAN) with a graph cutting method, the entire point cloud model is segmented at the component level to generate multiple unclassified component candidate regions. In each component region, two marker points with stable spatial distribution and significant geometric features are automatically extracted for subsequent component registration and skeleton splicing.

[0066] Attribute analysis layer: Extracts geometric information (size, boundary, curvature) of each component and combines it with spatially mapped material information to match its physical properties;

[0067] Component classification layer: A phased classification mechanism is used to prioritize the identification of load-bearing skeleton components (columns, beams, and walls), followed by functional components (doors, windows, and stairs).

[0068] Adjacency analysis layer: Constructs spatial adjacency graphs and structural connection graphs between components, uses graph neural networks to propagate structural connection relationships, and corrects preliminary classification results;

[0069] Classification output layer: Output the component category, geometric properties and physical properties to the component property database to form a structured component information table.

[0070] S203, component recognition model training:

[0071] The initial model is pre-trained using a few-shot learning strategy, and a training set is constructed using labeled building data (including labels for skeletal and functional components). A three-dimensional convolutional neural network (SparseConvNet) is used in conjunction with label-driven pre-training, and knowledge distillation is used to enhance small-sample recognition capabilities. A self-training mechanism is introduced, using unlabeled data to generate pseudo-labels and combining confidence assessment to screen high-confidence samples for incremental model training. During the training process, material information and component physical properties are introduced as auxiliary features to enhance the ability to distinguish classification boundaries and improve the accuracy of distinguishing highly similar components (load-bearing walls and partition walls). Combined with cross-validation and spatial structure consistency analysis, the recognition model undergoes multiple rounds of performance evaluation and parameter optimization, ultimately obtaining a stable component recognition model.

[0072] S204, component attribute database is established:

[0073] The spatial coordinates, category labels, geometric dimensions, physical properties, and registration markers of each identified component are stored in a structured manner; the database supports retrieval and query based on component category, spatial position, and attribute parameters, and provides interface services for subsequent bone structure splicing, performance analysis, and BIM modeling module calls.

[0074] Step S3: Analyzing the physical properties of the bone components in combination with the corresponding material information, and splicing the bone components into a bone structure model based on the analyzed physical property results;

[0075] S301, Extraction of bone component geometric properties and material information:

[0076] Extract the geometric properties (length, cross-sectional shape, thickness) and corresponding material information (material type and physical parameters) of all bone components classified as bone from the component attribute database generated in step S2. For each bone component, determine the corresponding material mechanical model (elastic model, plastic model) based on its geometric characteristics, and extract the material's mechanical parameters, including at least elastic modulus, yield strength, density, and thermal conductivity.

[0077] S302, Analysis of Physical Properties of Bone Components:

[0078] Based on the extracted geometric properties and material parameters, the finite element method is used to analyze the physical properties of each bone component. The analysis includes:

[0079] Load-bearing capacity: Evaluate the maximum load-bearing capacity of each bone component and analyze the mechanical response of the component under static and dynamic loads.

[0080] Bending stiffness: Calculates the bending stiffness of bone components under bending loads to assess their stability in the structure.

[0081] Node connection performance: Analyze the strength and stability of the connection nodes between bone components and other components, especially the connection performance under shear force and bending moment.

[0082] Material strength index: Combined with the material's tensile and compressive strength indicators, anti-destructive analysis is carried out.

[0083] Perform individual physical analysis on each bone component to obtain multi-dimensional performance data such as bearing capacity, stiffness, and connection strength, and generate a performance evaluation report.

[0084] S303, Modeling and Optimization of Bone Component Connection Relationships:

[0085] After analyzing the individual properties of each bone component, the connections between the components are rationally modeled based on their spatial relationships (relative position and geometry) and mechanical matching (coordinated mechanical properties and force patterns). A bone component connection map is constructed, where each node represents a bone component and edges represent the connection relationships between components. Edge weights represent the connection strength and coupling. Edge weights in the connection map are set based on the mechanical matching relationship (force transfer path and node strength) between each pair of bone components to ensure mechanical consistency. By combining spatial geometric information with mechanical properties, the spatial matching relationship between components is optimized, ensuring optimal stability of the overall structure under load.

[0086] S304: Based on the bone component connection atlas and building structure design specifications, a graph traversal algorithm (A* search algorithm) is used to topologically splice the bone components to generate a preliminary bone structure model. During the splicing process, the mechanical consistency and stability of all component connections are ensured. A finite element analysis is performed on the preliminarily spliced ​​bone structure model to evaluate its resistance and stability under various loading conditions. The analysis includes: overall stability and deformation analysis of the bone structure; response analysis of each bone component under different loads; and overall bending, shear, and torsion resistance of the structure. For structural weaknesses identified in the finite element analysis (insufficient load-bearing capacity, unreasonable node connections), local optimization adjustments are performed, including: redesigning local node connection schemes; optimizing the cross-sectional dimensions or material selection of bone components; and adjusting the spatial coordination of bone components to ensure that the overall structure meets performance requirements.

[0087] S305 generates a skeletal model that meets physical performance requirements based on the results of topological splicing and local optimization adjustments. This skeletal model serves as the foundation for subsequent BIM modeling and functional component integration. The final skeletal model data (including at least each component's geometry, mechanical properties, and connection relationships) is exported to a standardized format for easy use in subsequent steps. An API is provided to support data exchange and integration with other building design systems (such as structural analysis software and BIM platforms).

[0088] Step S4: aligning and replacing the bone structure model and the functional components in the three-dimensional spatial data model; calling a preset BIM component family template, automatically constructing the three-dimensional spatial data model after data replacement into a BIM model, and marking the structure corresponding to the bone structure model in the BIM model as the bone structure of the BIM model; the process of automatically constructing the BIM model complies with architectural design standards and specifications;

[0089] S401, spatial registration of bone structure model and functional components:

[0090] The generated bone structure model and functional components are integrated into a unified 3D spatial data model. The bone structure model serves as the primary framework, and the functional components are registered with the bone structure model based on their spatial position and geometric properties. The ICP algorithm (Iterative-Closest-Point) is used for precise registration of the point cloud and geometric model, ensuring geometric consistency between the bone structure and functional components in 3D space. For complex building components (curved staircases, irregular walls), a registration algorithm based on feature point matching is used, utilizing extracted landmarks for high-precision registration.

[0091] The least squares method is used to optimize the registration results to ensure accurate alignment of all component connections. If there are any geometric errors (deformation or misalignment) between the bone structure model and the functional components, these errors are adjusted through optimized mesh alignment until the optimal registration is achieved.

[0092] S402, Data Replacement and BIM Model Construction:

[0093] Based on the geometric properties, physical properties, and construction process information of each functional component, the preset BIM component family template is called, and the functional component data is replaced with standardized BIM component data that meets the template requirements. Each BIM component family template contains the geometric information, physical properties, and construction process information (installation method, construction process requirements) of the building component. The replaced three-dimensional spatial data model is converted into a BIM model, and an automated construction process is executed. This process includes: automatically generating corresponding component instances (beams, columns, walls, windows, doors) in the BIM based on the replaced component data, and automatically setting their position, size, and connection method; the skeletal structure model will serve as the skeletal structure in the BIM model (based on the geometric position and mechanical properties of the component), marked as a "structural part", and serve as the basic framework of the BIM model to support subsequent functional optimization and adjustment.

[0094] When generating a BIM model, the size, shape, position and connection method of the components are automatically adjusted based on the building design standards and specifications to ensure that the geometry of all components complies with the building specifications (spatial spacing, seismic requirements, construction interfaces).

[0095] S403, Integration and labeling of bone structure model and BIM model:

[0096] In the BIM model, the structural part corresponding to the skeletal structure model is specially marked and treated as a subset of the skeletal structure. Ensure that all structural components and functional components in the BIM model can be accurately mapped to the physical building. For different functional blocks, use color coding, layer division or attribute labels for visual distinction to ensure that the skeletal structure and functional components are clearly visualized in the BIM model. The components (columns, beams) and functional components (walls, windows) in the skeletal structure model are marked with clear connection points and relationships to ensure that the connection between the two meets structural safety requirements. Use the structural connection tool in the BIM platform (the connection module in Autodesk-Revit) to automatically establish the association relationship between the skeletal components and functional components to generate a complete building structure network.

[0097] S404, Quality Inspection and Compliance Assessment of BIM Models:

[0098] Automatically generated BIM models are checked for compliance with design specifications to ensure they meet architectural design standards (e.g., building space standards, mechanical performance standards, fire safety standards, etc.). Key checkpoints include: The rationality of structural member dimensions, node connection methods, and load transfer paths; the alignment of functional component locations, spacing, and dimensions with architectural drawing requirements; and the presence of conflicts (e.g., conflicts between pipes and beams, spatial overlap, etc.).

[0099] S405, BIM model output and subsequent use:

[0100] After the final BIM model is generated, it is exported to a standard BIM format to facilitate data exchange with other building design and construction management systems. The generated BIM model can be directly used for resource management, construction scheduling, construction team coordination, and other management tasks during the construction phase, enabling digital management of the building process. After the building is completed, the BIM model is also used for post-operation and maintenance management tasks such as facility management, equipment maintenance, and energy efficiency analysis, ensuring sustainable management throughout the building's lifecycle.

[0101] Step S5: Combine functional components to obtain design requirements, perform component-level adjustments on the existing BIM model based on preset design rules and inverse algorithms, and generate physical performance requirements for the BIM model. After the adjustments, perform physical performance analysis on the skeleton structure to ensure that it meets the physical performance requirements of the BIM model.

[0102] S501, obtain design requirements and functional component input:

[0103] Design objectives and requirements are input from the project owner, design team, or user. These requirements include the building's functional use, load-bearing capacity, energy efficiency, environmental friendliness, daylighting requirements, and fire resistance. Design objectives are detailed through functional requirements (room functions, window area, number of staircases), performance requirements (structural strength, seismic resistance, thermal comfort), and regulatory requirements (building safety standards, energy efficiency standards).

[0104] The geometric and physical properties of all functional components, as well as their spatial locations, are extracted from the BIM model. This input is then matched against a library of architectural design standards to ensure compliance with relevant design specifications (Architectural Design Code, Energy Efficiency Standards).

[0105] S502, component-level adjustment based on design rules and inverse algorithm:

[0106] A pre-set design rule library contains common architectural design rules (spatial layout, mechanical properties, energy efficiency requirements) and adjustment specifications for various building components (size, location, material selection). The design rule library also includes optimization guidelines based on experience or case studies to help designers optimize components based on project characteristics.

[0107] Using a goal-driven inverse algorithm, the design parameters for each component are derived based on user-entered design requirements. These parameters can include geometric adjustments (wall thickness, door and window dimensions, stair slope), physical property adjustments (material selection, load-bearing capacity, fire resistance), and spatial adjustments (such as room layout, door and window placement, and lighting and ventilation requirements). The inverse algorithm establishes an optimization function and combines it with a multi-objective optimization strategy (particle swarm optimization) to optimally configure building components. Optimization objectives include ensuring structural safety and stability; improving the building's energy efficiency, comfort, and environmental adaptability; and meeting the building's functional requirements (space utilization, daylighting).

[0108] S503, Generation and Correction of Physical Performance Requirements:

[0109] Combining all functional requirements and design rules, generate BIM model physical performance requirements that meet actual design objectives. Verify the physical performance of the skeleton structure (especially load-bearing components) to ensure it meets the overall building physical performance requirements. These physical performance requirements include: load-bearing capacity requirements, ensuring that the load capacity of each component in the structure meets expectations; seismic and wind resistance, designing the structure accordingly based on the natural environmental conditions (such as earthquakes and wind forces) of the building's location; thermal and energy efficiency requirements, ensuring that the building's thermal performance (such as insulation and ventilation) meets requirements based on energy-saving standards and design specifications; and noise control and comfort, ensuring that the building provides a comfortable noise environment during use through adjustments to sound insulation design.

[0110] Based on these physical performance requirements, the geometry and material properties of the skeleton and other related components are adjusted to ensure that the model meets the comprehensive physical performance requirements of the building. Verification is performed during the adjustment process to ensure that each adjustment enhances the feasibility of the model while adhering to the basic principles of architectural design.

[0111] S504, Bone Structure Physical Performance Analysis and Verification:

[0112] After adjustment, the physical properties of the bone structure are analyzed, including:

[0113] Load-bearing capacity analysis: Through structural mechanics simulation, the response of the adjusted bone components under load is evaluated to ensure that overloading or damage does not occur;

[0114] Node and connection performance: Check the stability of each connection node of the bone structure to confirm that the connection points can safely and effectively transfer loads;

[0115] Overall stability: Evaluate the overall stability of the entire bone structure to ensure that its earthquake resistance, wind resistance and other capabilities meet the design standards;

[0116] Energy efficiency and thermal comfort analysis: Conduct thermal and energy efficiency analysis on the skeleton structure and its connections with other components to ensure that the building structure meets energy conservation and emission reduction targets.

[0117] After completing the physical performance analysis, the final BIM model is confirmed to meet all design requirements and is updated and archived accordingly. This version of the BIM model will serve as the basic data for building construction and operation and maintenance.

[0118] Step S6: Performing a comprehensive performance analysis on the adjusted BIM model, including structural mechanical performance, energy consumption simulation, and natural lighting; matching optimization suggestions with corresponding design requirements based on the comprehensive performance analysis results to drive automatic iterative updates of the BIM model;

[0119] S601, comprehensive performance analysis model construction:

[0120] Construct a comprehensive performance analysis framework to convert the building's various performance requirements (structure, energy efficiency, and lighting) into quantifiable analysis parameters.

[0121] The framework includes:

[0122] Structural mechanical properties - assess the stability and load-bearing capacity of building structures;

[0123] Energy efficiency simulation - analysis of building energy efficiency, thermal performance and energy consumption performance;

[0124] Daylight Analysis - Evaluate the amount of natural light inside a building to ensure comfort and energy efficiency goals are met.

[0125] S602, Structural Mechanical Performance Analysis:

[0126] Use structural mechanics analysis tools (such as SAP2000) to conduct a comprehensive structural mechanics performance analysis of the adjusted BIM model. The main analysis contents include:

[0127] Bearing capacity: Analyze the bearing capacity of each component under different loads to ensure that problems such as yielding and fracture will not occur;

[0128] Seismic performance: Considering the earthquake intensity of the area where the building is located, analyze the building's response under earthquake loads to ensure that the building meets seismic design requirements;

[0129] Node connection and mechanical stability: By analyzing the connection nodes, the connection method and node stability between building components are evaluated to ensure the stability of the overall structure;

[0130] Local effects: Conduct local stress analysis on large-span components and special components (such as the facade structure of high-rise buildings) to ensure that they will not become unstable under extreme load conditions.

[0131] Structural analysis results provide information such as stress, strain, and displacement to help identify potential structural safety hazards and provide optimization recommendations based on these results. Based on the analysis results, components requiring adjustment can be identified and sized, material selected, or enhanced to improve structural safety and stability.

[0132] S603, Energy Efficiency Simulation and Optimization:

[0133] EnergyPlus is used to simulate the building's energy efficiency and analyze the building's thermal comfort, energy consumption, heating and cooling loads, and air conditioning requirements. Specific analysis includes:

[0134] Heat transfer analysis: Calculate the heat transfer performance of building exterior walls, roofs, windows, etc., and analyze the building's thermal insulation effect;

[0135] Energy efficiency assessment: Evaluate the overall energy efficiency level based on the building's structural layout, lighting, ventilation, insulation, etc., and evaluate the building's heating and cooling loads according to different seasons and climate conditions;

[0136] Renewable energy utilization: Analyze the building's use of renewable energy (such as solar energy, wind energy, geothermal energy, etc.) and evaluate its contribution to the building's energy efficiency.

[0137] Based on the energy efficiency simulation results and in conjunction with energy-saving design specifications, optimization recommendations are provided, such as adjusting the thermal insulation layer of the building envelope, optimizing window orientation, and adding solar water heaters.

[0138] S604, Natural Lighting Analysis and Optimization:

[0139] The natural lighting inside the building is analyzed using the lighting simulation tool (Dialux) to ensure that the building's lighting system can maximize the use of natural light. The main analysis contents include:

[0140] Sunlight duration and intensity: Analyze the natural lighting conditions in various areas of the building at different times, especially in important functional areas (such as office areas and living rooms);

[0141] Lighting ratio and uniformity: Evaluate the uniformity of indoor lighting to avoid uneven lighting or dark areas;

[0142] Combining lighting and energy efficiency: Combine lighting analysis with energy efficiency simulation to evaluate the contribution of lighting optimization to building energy efficiency.

[0143] Based on the results of the daylighting analysis, adjust the building's window size, orientation, and shading measures to improve daylighting efficiency and spatial comfort. For areas requiring increased daylighting, consider adding skylights or sun ducts. For areas with excessive daylighting, adjust with shading devices or curtains to reduce energy consumption and improve comfort.

[0144] S605, Comprehensive Performance Evaluation and Feedback Mechanism:

[0145] Integrate the results of structural mechanical performance, energy efficiency, and daylighting analyses into a comprehensive performance assessment report. This report provides the project team with a comprehensive performance assessment, helping decision-makers understand the building's performance across all aspects. Results are presented through charts, heat maps, simulation videos, and other formats, allowing designers and clients to intuitively understand the building's performance.

[0146] Based on the comprehensive performance analysis results, optimization suggestions are generated, including specific suggestions on structural optimization, energy efficiency improvement, and lighting improvement.

[0147] S606, Automatic iterative update and verification of BIM models:

[0148] Based on the comprehensive performance analysis results and optimization recommendations, the system automatically adjusts each component in the BIM model, modifying its geometric and physical properties and updating the adjusted BIM model to the latest version. This update process includes automated component replacement, sizing, and material selection to ensure the BIM model meets the optimized design requirements. The updated BIM model undergoes further structural mechanics analysis, energy efficiency simulation, and daylighting analysis to ensure that the adjusted model meets all performance requirements. This verification process ensures that all building design requirements are met, as well as building codes and project requirements.

[0149] Step S7: Support users to dynamically adjust design requirements, display various BIM models and their comprehensive performance analysis results obtained during the optimization process through a visual interface, realize rapid "what-if" analysis and scheme comparison, and help users make decisions among multiple design schemes.

[0150] Visualize the results of comprehensive performance analysis and present the optimized building design through charts, heat maps, 3D simulations, and other formats, enabling the design team to intuitively understand the building's performance. Leveraging the BIM platform's visual interface to showcase the updated model, designers and clients can quickly compare and contrast different design options.

[0151] Dynamic adjustments are supported. Through "what-if" analysis, users can quickly compare multiple design options, view the impact of adjustments on building performance, and make the optimal decision among multiple options. Based on different optimization goals and design requirements, the system presents multiple design options for decision makers to choose from, and evaluates each option's overall performance in real time.

[0152] Example 2, a three-dimensional digital survey and design system based on BIM, see Figure 1As shown, it includes the following modules:

[0153] The data acquisition module obtains the point cloud data and corresponding material information of the target building and generates an initial three-dimensional spatial data model;

[0154] The component recognition module uses a component recognition model based on a three-dimensional spatial data model to automatically identify and classify the main components in the building, extract the geometric and physical properties of the components, and establish a component property database;

[0155] The physical performance analysis module analyzes the physical properties of bone components based on material information and assembles the bone components into a bone structure model based on the analysis results;

[0156] The BIM modeling module aligns and replaces the bone structure model and functional components in the three-dimensional spatial data model, calls the preset BIM component family template, automatically generates the BIM model, and obtains the bone structure;

[0157] The reverse adjustment module adjusts the existing BIM model at the component level based on preset design rules and reverse inference algorithms, generates the physical performance requirements of the BIM model, and analyzes the skeleton to ensure that it meets the physical performance requirements of the BIM model;

[0158] The performance analysis and optimization module conducts comprehensive performance analysis on the adjusted BIM model, matches optimization suggestions based on the analysis results, and drives automatic iterative updates of the BIM model;

[0159] The user interaction module supports users to dynamically adjust design requirements and displays various BIM models and their comprehensive performance analysis results during the optimization process through a visual interface.

[0160] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A three-dimensional digital survey and design method based on BIM, characterized in that: include: Obtain the point cloud data and corresponding material information of the target building, and perform cleaning, coordinate registration, and splicing to generate an initial 3D spatial data model; Based on the three-dimensional spatial data model, a component recognition model is used to automatically identify and classify the main components in the building, and the geometric and physical properties of the components are extracted to establish a component property database; the classification includes skeleton components and functional components; The component retains two markers for registration; Performing physical property analysis on the bone components in combination with the corresponding material information, and splicing the bone components into a bone structure model based on the physical property analysis results; The bone structure model comprises: The geometric properties and material information of all bone components are obtained, and the physical properties of each bone component are analyzed based on the finite element analysis method. The physical properties analysis includes load-bearing capacity, bending stiffness, node connection performance, and material strength indicators. Based on the analysis results, bone component pairs that meet structural safety and connection feasibility are selected, and a bone component connection map is constructed based on the spatial position relationship and mechanical matching relationship of the components. The edge weights in the connection map reflect the connection strength and component coupling degree. In combination with the component connection map and building structure design specifications, the bone components are topologically spliced ​​using a graph traversal algorithm and mechanical consistency constraints to generate an initial bone structure model. Based on the initial bone structure model, a finite element analysis of the entire structure is further performed to evaluate the overall stability and resistance performance of the spliced ​​structure, and the component connection relationships that do not meet the performance requirements are locally optimized and adjusted until a bone structure model that meets the target physical performance requirements is generated. The skeletal structure model and functional components are aligned and data replaced in the three-dimensional spatial data model; the preset BIM component family template is called to automatically construct the three-dimensional spatial data model after data replacement into a BIM model, and the architecture corresponding to the skeletal structure model in the BIM model is marked as the skeletal architecture of the BIM model; the process of automatically constructing the BIM model complies with architectural design standards and specifications; The constructing of the BIM model includes: automatically matching and replacing components in the model by calling a preset BIM component family template based on the replaced three-dimensional spatial data model, wherein the component family template includes the geometry, physical properties and construction process of the building components; automatically adjusting the size, position and connection method of the components according to the architectural design standards and specifications, and generating a complete BIM model, and marking the corresponding content of the bone structure model in the BIM model as the bone structure; Combine functional components to obtain design requirements. Based on preset design rules and reverse engineering algorithms, adjust the existing BIM model at the component level and generate the physical performance requirements of the BIM model. After the adjustments, perform physical performance analysis on the skeleton structure to ensure that it meets the physical performance requirements of the BIM model. The component-level adjustment includes: The system obtains the design goals input by the user, calls the goal-driven rules and component constraint rules in the design rule library, establishes an optimization function based on the multi-objective optimization algorithm, and generates component-level change suggestions that meet the design goals through reverse calculation. The component adjustment process is combined with the performance impact assessment mechanism to evaluate the impact of component changes on structural performance, energy consumption performance or lighting efficiency in real time, and dynamically adjust the optimization strategy accordingly.

2. A BIM-based three-dimensional digital survey and design method according to claim 1, characterized in that: Also includes: Performing a comprehensive performance analysis on the adjusted BIM model, including structural mechanical performance, energy consumption simulation, and natural lighting; Match optimization suggestions and corresponding design requirements based on comprehensive performance analysis results to drive automatic iterative updates of the BIM model; It supports users to dynamically adjust design requirements, and displays various BIM models and their comprehensive performance analysis results obtained during the optimization process through a visual interface, enabling rapid simulation hypothesis analysis and scheme comparison, helping users make decisions among multiple design schemes.

3. The BIM-based three-dimensional digital survey and design method according to claim 1, characterized in that: The structure of the component identification model includes: Input layer, used to obtain the three-dimensional spatial data model of the building; Feature extraction layer, used to extract multi-scale spatial features of point cloud data in the three-dimensional spatial data model and construct a feature pyramid; The component segmentation layer is used to segment the entire building according to the feature pyramid, generate multiple unclassified components, and mark each component with two spatial marker points for subsequent alignment; The attribute analysis layer is used to extract the geometric properties of each component and obtain the physical properties of the component in combination with the corresponding material information; The component classification layer is used to perform functional analysis and classification of unclassified components. It adopts a staged classification method, first identifying all bone components and then identifying functional components; The adjacency association analysis layer is used to analyze the spatial adjacency and structural connection relationships between components, and to correct the results of the component classification layer based on the association information; The classification output layer is used to output the final identified component category, geometric properties and physical properties to the component property database.

4. A BIM-based three-dimensional digital survey and design method according to claim 3, characterized in that: The training of the component recognition model includes: A three-dimensional spatial data model of a building with complete annotated skeletal and functional components is obtained as a data sample, and an initial training set is constructed. The three-dimensional convolutional neural network is pre-trained using a few-shot learning method combined with the initial training set to obtain an initial component recognition model. A self-training mechanism is used to perform incremental training in combination with a large number of unannotated three-dimensional spatial data models of buildings. The parameters of the initial component recognition model are optimized through pseudo-label generation and confidence screening mechanisms to improve the recognition accuracy of different types of components. Component physical properties and material information are introduced as auxiliary features during the training process to optimize the discrimination boundary of component classification and improve the distinction between functional and skeletal components. The training results are evaluated for accuracy through cross-validation and structural adjacent relationship analysis, and the optimal model parameter set for component recognition is finally determined to obtain the final component recognition model.

5. A three-dimensional digital survey and design system based on BIM, characterized in that: The system applies a BIM-based three-dimensional digital survey and design method according to any one of claims 1 to 4, comprising: The data acquisition module obtains the point cloud data and corresponding material information of the target building and generates an initial three-dimensional spatial data model; The component recognition module uses a component recognition model based on a three-dimensional spatial data model to automatically identify and classify the main components in the building, extract the geometric and physical properties of the components, and establish a component property database; The physical performance analysis module analyzes the physical properties of bone components based on material information and assembles the bone components into a bone structure model based on the analysis results; The BIM modeling module aligns and replaces the bone structure model and functional components in the three-dimensional spatial data model, calls the preset BIM component family template, automatically generates the BIM model, and obtains the bone structure; The reverse adjustment module adjusts the existing BIM model at the component level based on preset design rules and reverse inference algorithms, generates the physical performance requirements of the BIM model, and analyzes the skeleton to ensure that it meets the physical performance requirements of the BIM model; The performance analysis and optimization module conducts comprehensive performance analysis on the adjusted BIM model, matches optimization suggestions based on the analysis results, and drives automatic iterative updates of the BIM model; The user interaction module supports users to dynamically adjust design requirements and displays various BIM models and their comprehensive performance analysis results during the optimization process through a visual interface.

Images