Three-dimensional digital survey design method and system based on BIM (Building Information Modeling)

Abstract

The invention discloses a BIM-based three-dimensional digital survey design method and system, and relates to the technical field of building survey design. A three-dimensional digital survey design system based on BIM comprises a data acquisition module, a component identification 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. According to the method, main components in a building are automatically identified and classified through a component identification model, geometric attributes and physical attributes of the components are extracted, a component attribute database is established, the model has the bone component and functional component classification capacity, mark points needed by registration are reserved, accurate mapping from point cloud to structural semantics is achieved, and the accuracy of the component identification model is improved. And the artificial modeling burden is obviously reduced.

Description

Technical Field

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

[0002] With the in-depth transformation of the construction industry towards digitalization and intelligentization, the Building Information Modeling (BIM) technology has been increasingly widely applied in the stages of building design, construction, and operation. The BIM technology can integrate the three-dimensional spatial information, component attributes, construction processes, and management requirements of a building into a complete information model, providing a comprehensive and visual digital platform for building design, construction, and operation. However, in practical applications, there are still some problems with the design methods based on traditional manual surveying and two-dimensional drawings, especially facing great challenges in aspects such as building design accuracy and design efficiency.

[0003] Traditional architectural survey methods mainly rely on manual measurement, two-dimensional drawings, and manual modeling. These methods are not only easily limited by human errors but also difficult to quickly and accurately obtain the true three-dimensional data of a building. In addition, existing architectural survey and design solutions usually lack sufficient digitalization and intelligentization, making it difficult to effectively evaluate and optimize various complex factors (such as structural safety, energy consumption, daylighting, etc.) in the building design process.

[0004] On this basis, in recent years, three-dimensional digital design methods based on BIM technology have emerged. By combining building point cloud data with the BIM model, full-life cycle management from survey, design to construction can be achieved. However, existing BIM-based design methods still have several limitations, mainly including low automation, insufficient component recognition accuracy, lack of structural analysis for model optimization, and 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 obtains components by combining a three-dimensional spatial data model, generates a bone structure model, jointly uses the automatically generated BIM model, and reasonably optimizes itself through the bone framework and backstepping algorithm of the BIM model to solve the above problems.

[0006] A three-dimensional digital survey and design method based on BIM includes: Obtaining the point cloud data and corresponding material information of the target building, and performing cleaning, coordinate registration, and splicing processing to generate an initial three-dimensional spatial data model; Based on a three-dimensional spatial data model, a component recognition model is used to automatically recognize and classify the main components in a building, extract the geometric and physical properties of the components, and establish a component property database; the classification includes bone components and functional components; two marker points are reserved for each component for registration; Combined with the corresponding material information, perform physical property analysis on the bone components, and according to the analysis results of the physical properties, splice the bone components into a bone structure model; Register and replace the data of the bone structure model and the functional components in the three-dimensional spatial data model; call the preset BIM component family template, and automatically construct the three-dimensional spatial data model after data replacement into a BIM model, and mark the corresponding architecture of the bone structure model in the BIM model as the bone architecture of the BIM model; the process of automatically constructing into a BIM model follows the architectural design standards and specifications; Combined with the functional components to obtain the design requirements, based on the preset design rules and reverse inference algorithms, perform component-level adjustment on the existing BIM model, and generate the physical property requirements of the BIM model. After adjustment, perform physical property analysis on the bone architecture to ensure compliance with the physical property requirements of the BIM model.

[0007] As a preferred technical solution of the present invention, a three-dimensional digital survey and design method based on BIM further includes: Perform comprehensive performance analysis on the adjusted BIM model, and the performance analysis includes structural mechanics performance, energy consumption simulation, and natural lighting; match optimization suggestions and corresponding design requirements according to the comprehensive performance analysis results to drive the automatic iterative update of the BIM model; Support users to dynamically adjust the design requirements, display each BIM model and its 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.

[0008] As a preferred technical solution of the present invention, the structure of the component recognition model includes: An input layer for obtaining the three-dimensional spatial data model of the building; A feature extraction layer for extracting the multi-scale spatial features of the point cloud data in the three-dimensional spatial data model and constructing a feature pyramid; A component segmentation layer for segmenting the whole building according to the feature pyramid, generating multiple unclassified components, and marking two spatial marker points for subsequent registration for each component; An attribute analysis layer for extracting the geometric attributes of each component and obtaining the physical attributes of the component in combination with the corresponding material information; A component classification layer for performing functional analysis and classification on the unclassified components, adopting a phased classification method, first identifying all bone components, and then identifying functional components; An adjacent relationship analysis layer for analyzing the spatial adjacency relationship and structural connection relationship between components, and correcting the results of the component classification layer based on the association information; A classification output layer for outputting the finally identified component categories, geometric attributes, and physical attributes to the component attribute database.

[0009] As a preferred technical solution of the present invention, the training of the component recognition model includes: Obtaining a three-dimensional spatial data model of a building with fully labeled bone components and functional components as a data sample, constructing an initial training set, and pre-training a three-dimensional convolutional neural network using the few-shot learning method in combination with the initial training set to obtain an initial component recognition model; adopting a self-training mechanism, and performing incremental training in combination with the three-dimensional spatial data models of a large number of unlabeled buildings, and optimizing the parameters of the initial component recognition model through a pseudo-label generation and confidence screening mechanism to improve the recognition accuracy of different types of components; introducing the physical attributes and material information of components as auxiliary features during the training process to optimize the discrimination boundary of component classification and improve the distinguishability between functional components and bone components; evaluating the accuracy of the training results through cross-validation and structural adjacency relationship analysis, and finally determining the optimal model parameter set for component recognition to obtain the final component recognition model.

[0010] As a preferred technical solution of the present invention, the bone structure model includes: Obtaining the geometric attributes and material information of all bone components, and performing single-component physical property analysis on each bone component based on the finite element analysis method. The physical property analysis includes bearing capacity, flexural stiffness, joint connection performance, and material strength index; according to the analysis results, screening bone component pairs that meet the structural safety and connection feasibility, and constructing a bone component connection map based on the spatial position relationship and mechanical matching relationship of the components. The edge weight in the connection map reflects the connection strength and component coupling degree; combining the component connection map with the building structure design specifications, and using a graph traversal algorithm and mechanical consistency constraints to perform topological splicing on the bone components to generate an initial bone structure model; on the basis of the initial bone structure model, further performing overall structure finite element analysis to evaluate the overall stability and resistance performance of the spliced structure, and locally optimizing and adjusting the component connection relationship that does not meet the performance requirements until a bone structure model that meets the target physical property requirements is generated.

[0011] As a preferred technical solution of the present invention, the construction of the BIM model includes: based on the replaced three-dimensional spatial data model, by calling a preset BIM component family template, automatically matching and replacing the components in the model, and the component family template includes the geometry, physical properties and construction techniques of building components; automatically adjusting the component dimensions, positions and connection methods according to the building design standards and specifications, generating a complete BIM model, and marking the corresponding content of the bone structure model in the BIM model as the bone framework.

[0012] As a preferred technical solution of the present invention, the component-level adjustment includes: Obtaining the design objectives input by the user, calling the target-driven rules and component constraint rules in the design rule library, establishing an optimization function in combination with a multi-objective optimization algorithm, and generating component-level change suggestions that meet the design objectives through back-calculation; the component adjustment process combines a performance impact assessment mechanism to evaluate the impact of component changes on structural performance, energy consumption performance or daylighting efficiency in real time, and dynamically adjusts the optimization strategy accordingly.

[0013] A three-dimensional digital survey and design system based on BIM includes: A data acquisition module that obtains the point cloud data of the target building and the corresponding material information, and generates an initial three-dimensional spatial data model; A component recognition module that, based on the three-dimensional spatial data model, automatically recognizes and classifies the main components in the building using a component recognition model, extracts the geometric and physical properties of the components, and establishes a component attribute database; A physical performance analysis module that performs physical performance analysis on the bone components in combination with the material information, and splices the bone components into a bone structure model according to the analysis results; A BIM modeling module that registers and replaces the bone structure model and functional components in the three-dimensional spatial data model, calls a preset BIM component family template, automatically generates a BIM model, and obtains the bone framework; A reverse adjustment module that, based on preset design rules and back-calculation algorithms, performs component-level adjustment on the existing BIM model, generates the physical performance requirements of the BIM model, and analyzes the bone framework to ensure that it meets the physical performance requirements of the BIM model; A performance analysis and optimization module that performs comprehensive performance analysis on the adjusted BIM model, matches optimization suggestions according to the analysis results, and drives the automatic iterative update of the BIM model; A user interaction module that supports the user to dynamically adjust the design requirements, and displays each BIM model and its comprehensive performance analysis results during the optimization process through a visual interface.

[0014] The present invention has the following advantages: The present invention automatically identifies and classifies the main components in a building through a component recognition model, extracts the geometric and physical attributes of the components, and establishes a component attribute database. This model has the ability to classify bone components and functional components, and retains the marking points required for registration, realizing an accurate mapping from point cloud to structural semantics, and significantly reducing the burden of manual modeling.

[0015] Through physical performance analysis, the present invention assembles bone components to form a bone structure model, which serves as the "bone framework" for subsequent BIM model construction. This method realizes a BIM modeling process driven by structural performance, ensures the structural safety and performance controllability of the building model, and achieves an integrated modeling of "form-structure-performance".

[0016] Based on the registration of the bone structure model and functional components, the present invention calls a preset BIM component family template to realize automatic replacement of components and standardized reconstruction of the model structure, and automatically generates a BIM model that complies with building codes, significantly improving the modeling efficiency and standard consistency.

[0017] The present invention supports combining functional components with user design requirements, and based on preset design rules and reverse deduction algorithms, making component-level adjustments to the existing BIM model. This process not only considers the design objectives, but also combines physical performance requirements for optimization analysis, realizing an intelligent feedback adjustment from performance to form; a comprehensive performance analysis will be carried out on the adjusted BIM model, and optimization suggestions will be generated according to the results, driving the automatic iterative update of the BIM model, and greatly improving the scientificity and rationality of building design. Brief Description of the Drawings

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for description in the embodiments or the prior art. Obviously, the drawings in the following description are only schematic diagrams of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to the provided drawings; Figure 1 It is a structural schematic diagram of a 3D digital survey and design system based on BIM adopted in an embodiment of the present invention. Detailed Embodiments

[0019] In order to make the objectives, technical solutions and advantages of the present invention clearer, the following will further describe the present invention in detail with reference to the drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0020] Embodiment 1, a 3D digital survey and design method based on BIM, includes the following steps: 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 3D spatial data model; S101 Point cloud data acquisition: Use a laser scanning device (SLAM system) or a 3D scanner carried by a drone to perform panoramic scanning of the target building to obtain the original point cloud data covering the entire structure of the building. During the scanning process, at least record the spatial coordinates, reflection intensity, and RGB color values of each acquisition point, and classify them according to the scanning stations. Combine manual sampling records, building drawing analysis, and IoT sensing devices to collect and label the material types, material numbers, and mechanical property parameters (density, elastic modulus, thermal conductivity) of each area of the building, construct a material attribute information table, and correspond and match it with the point cloud data according to the spatial area.

[0021] S102, preprocess the original point cloud data, including: Noise removal (isolated point filtering, voxel filtering); Data compression and resampling (downsampling processing based on voxel grids); Identification and removal of invalid points and duplicate points; Remove non-building-related background data (ground debris, moving targets).

[0022] S103, coordinate registration processing: In the case of multi-station scanning or multi-view acquisition, use a geometric registration algorithm based on marker points to spatially align the point clouds of each sub-region; splice and fuse the spatially registered point cloud data of each region to generate a complete building point cloud model; divide the large-scale point cloud data into multiple data blocks according to the building area or functional area to improve the subsequent processing efficiency; store the point cloud data structurally to form an initial 3D spatial data model that can be used for subsequent component recognition.

[0023] Output the structured format (.pcd) of the 3D spatial data model; perform spatial mapping and association between the point cloud data and the material attribute data to provide basic data support for subsequent component recognition and physical property analysis.

[0024] Step S2: Based on the 3D spatial data model, use a component recognition model to automatically identify and classify the main components in the building, extract the geometric and physical attributes of the components, and establish a component attribute database; the classification includes bone components and functional components; two marker points are reserved for each component for registration; S201, component recognition model construction and input preprocessing: Based on the three-dimensional spatial data model output in step S1, input it into a preset component recognition model; perform normalization processing on the point cloud data, and adopt a hierarchical sampling strategy to construct data views with different density levels respectively; divide the point cloud into processable sub-regions according to spatial blocks, adapt to the input dimension of the neural network model, and construct a multi-scale feature input structure.

[0025] S202, Structure of the component recognition model: The component recognition model is a multi-module recognition system based on the fusion architecture of a three-dimensional convolutional neural network (3D-CNN) and a graph neural network (GNN), specifically including: Input layer: Receive the normalized three-dimensional spatial data model; Feature extraction layer: Extract local and global spatial features of the point cloud through voxelization processing and the PointNet++ network, and construct a feature pyramid; Component segmentation layer: Combine the spatial clustering algorithm (DBSCAN) and the graph cut method to perform component-level regional segmentation on the overall point cloud model, generating multiple unclassified component candidate regions; in each component region, automatically extract two marker points with stable spatial distribution and significant geometric features for subsequent component registration and skeleton splicing; Attribute analysis layer: Extract its geometric shape information (size, boundary, curvature) for each component, and combine the spatial mapping material information to match its physical attributes; Component classification layer: Adopt a phased classification mechanism to first identify load-bearing bone components (columns, beams, walls), and then identify functional components (doors, windows, stairs); Adjacent connection analysis layer: Construct a spatial adjacency graph and a structural connection graph between components, and use the graph neural network to propagate the structural connection relationship to correct the preliminary classification results; Classification output layer: Output the category, geometric attributes, and physical attributes of the component to the component attribute database to form a structured component information table.

[0026] S203, Training of the component recognition model: The initial model is pre-trained through a few-shot learning strategy, and a training set is constructed using labeled building data (including labels of bone components and functional components). A three-dimensional convolutional neural network (SparseConvNet) is used for pre-training in combination with label driving, and the knowledge distillation method is utilized to improve the few-shot recognition ability. A self-training mechanism is introduced, where unlabeled data is used to generate pseudo-labels and high-confidence samples are screened by combining confidence evaluation, and incremental model training is performed. During the training process, material information and component physical properties are introduced as auxiliary features to enhance the classification boundary discrimination ability and improve the discrimination accuracy of highly similar components (load-bearing walls and partition walls). Combining cross-validation and spatial structure consistency analysis, multi-round performance evaluation and parameter optimization are carried out on the recognition model, and finally a stable component recognition model is obtained.

[0027] S204, Establishment of component attribute database: The spatial coordinates, category labels, geometric dimensions, physical properties, and registration marker points of each recognized component are stored in a structured manner; the database supports retrieval and query according to component categories, spatial positions, and attribute parameters, and provides interface services for subsequent bone structure splicing, performance analysis, and BIM modeling module calls.

[0028] Step S3: Perform physical property analysis on bone components in combination with corresponding material information, and splice the bone components into a bone structure model according to the analyzed physical property results; S301, Extraction of geometric attributes and material information of bone components: Extract the geometric attributes (length, cross-sectional shape, thickness) of all bone components classified as such and their corresponding material information (material type and its physical parameters) from the component attribute database generated in step S2. For each bone component, in combination with its geometric characteristics, determine the corresponding material mechanics model (elastic model, plastic model), and extract the mechanical parameters of the material, including at least elastic modulus, yield strength, density, and thermal conductivity.

[0029] S302, Physical property analysis of individual bone components: According to the extracted geometric attributes and material parameters, use the finite element method to perform physical property analysis on each individual bone component. The analysis content includes: 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.

[0030] Flexural stiffness: Calculate the flexural stiffness of the bone component under bending load and evaluate its stability in the structure.

[0031] Joint 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.

[0032] Material strength index: Combine the tensile and compressive strength and other indicators of the material to conduct anti-destruction analysis.

[0033] Conduct separate 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.

[0034] S303, Modeling and optimization of the connection relationship of bone components: After analyzing the single performance of each bone component, based on the spatial relationship (relative position, geometric shape) and mechanical matching (cooperating mechanical properties, force application mode) between bone components, reasonably model the connection between bone components. Construct a connection map of bone components, where each node represents a bone component and the edge represents the connection relationship between bone components. The weight of the edge represents the connection strength and coupling degree. Set the edge weights in the connection map based on the mechanical matching relationship (force transmission path, node strength) of each pair of bone component connections to ensure mechanical consistency. Combine spatial geometric information and mechanical characteristics to optimize the spatial cooperation relationship between components, so that the overall structure reaches the best stability under force.

[0035] S304, Based on the connection map of bone components and building structure design specifications, use the graph traversal algorithm (A* search algorithm) to perform topological splicing on bone components to generate a preliminary bone structure model. During the splicing process, ensure the mechanical consistency and stability of all component connections. Conduct overall finite element analysis on the preliminarily spliced bone structure model to evaluate its resistance performance and stability under various load conditions. The analysis content includes: overall stability and deformation analysis of the bone structure; response analysis of each bone component under different loadings; overall bending, shear, and torsion resistance capabilities of the structure. For the structurally weak areas (insufficient bearing capacity, unreasonable node connections) that appear in the finite element analysis, conduct local optimization and adjustment, including: redesigning the local node connection scheme; optimizing the cross-sectional dimensions or material selection of bone components; adjusting the spatial cooperation position of bone components to ensure that the overall structure meets the performance requirements.

[0036] S305 According to the results of topological splicing and local optimization and adjustment, finally generate a bone structure model that meets the physical performance requirements. This bone structure model will serve as the basis for subsequent BIM modeling and functional component integration. Output the data of the final bone structure model (including at least the geometric information, mechanical properties, and connection relationship of each component) in a standardized format for use in subsequent steps. Provide an API interface to support data interaction and integration with other building design systems (such as structural analysis software, BIM platform).

[0037] Step S4: Register and replace data for the bone structure model and functional components in the three-dimensional spatial data model; call the preset BIM component family template, and automatically construct the three-dimensional spatial data model after data replacement into a BIM model, and mark the architecture corresponding to the bone structure model in the BIM model as the bone architecture of the BIM model; the process of automatically constructing into a BIM model follows the architectural design standards and specifications. S401, Spatial registration of the bone structure model and functional components: Integrate the generated bone structure model and functional components into a unified three-dimensional spatial data model. At this time, the bone structure model serves as the main framework, and the functional components need to be registered with the bone structure model according to their spatial positions, geometric attributes. The Iterative-Closest-Point (ICP) algorithm is used for accurate registration of the point cloud and geometric model to ensure the geometric consistency of the bone structure and functional components in three-dimensional space; for complex building components (curved stairs, irregular walls), a registration algorithm based on feature point matching is used to perform high-precision registration using the extracted marker points.

[0038] Combine the least squares method to optimize the error of the registration result to ensure the precise docking of the connection nodes of all components. If there are mismatched geometric errors (deformation, dislocation) between the bone structure model and functional components, they are adjusted by optimizing the mesh alignment until the best registration effect is achieved.

[0039] S402, Data replacement and BIM model construction: Based on the geometric attributes, physical attributes, and construction process information of each functional component, call the preset BIM component family template, and replace the functional component data with standardized BIM component data that meets the template requirements. Each BIM component family template contains the geometric information, physical attributes, and construction process information (installation method, construction process requirements) of the building component. Convert the three-dimensional spatial data model after replacement into a BIM model and execute an automated construction process. This process includes: automatically generating the corresponding component instances (beams, columns, walls, windows, doors) in BIM according to the replaced component data, and automatically setting their positions, sizes, and connection methods; the bone structure model will serve as the bone architecture in the BIM model (based on the geometric position and mechanical properties of the components), marked as "structural part", and serve as the basic framework of the BIM model to support subsequent function optimization and adjustment.

[0040] When generating the BIM model, automatically adjust the size, shape, position, and connection method of the components based on the architectural design standards and specifications to ensure that the geometric forms of all components meet the building codes (spatial spacing, seismic requirements, construction interfaces).

[0041] S403, Integration and marking of the bone structure model and BIM model: In the BIM model, specifically mark the architectural part corresponding to the bone structure model and use it as a subset of the bone architecture part. 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 tags for visual differentiation to ensure clear visualization of the bone architecture part and functional components in the BIM model. The components (columns, beams) in the bone structure model and the functional components (walls, windows) are labeled with clear connection points and relationships to ensure that the connection method between the two meets the requirements of structural safety. Use the structured connection tool in the BIM platform (the connection module in Autodesk - Revit) to automatically establish the association relationship between the bone components and the functional components and generate a complete building structure network.

[0042] S404, Quality Inspection and Compliance Assessment of the BIM Model: Conduct a compliance check of the design specifications on the automatically generated BIM model to ensure that the model meets the building design standards (such as building space standards, mechanical performance standards, fire safety standards, etc.). The main inspection points include: whether the dimensions of the bone components, the connection methods of the nodes, and the load transfer paths are reasonable; whether the positions, spacings, and dimensions of the functional components match the requirements of the building drawings; whether there are conflicts (such as conflicts between pipes and beams / columns, spatial overlaps, etc.).

[0043] S405, BIM Model Output and Subsequent Utilization: After the final BIM model is generated, it is output in the standard BIM format for data interaction with other building design and construction management systems. The generated BIM model can be directly used for management work such as resource management, construction schedule arrangement, and construction team coordination during the construction phase to achieve digital management of the building construction process. After the building is completed, the BIM model is also used for post - operation and maintenance management work such as facility management, equipment maintenance, and energy efficiency analysis to ensure sustainable management throughout the building's life cycle.

[0044] Step S5: Obtain the design requirements in combination with the functional components, based on the preset design rules and reverse - deduction algorithms, make component - level adjustments to the existing BIM model, and generate the physical performance requirements of the BIM model. After the adjustment, conduct a physical performance analysis of the bone architecture to ensure compliance with the physical performance requirements of the BIM model.

[0045] S501, Obtain Design Requirements and Input of Functional Components: Design goals and requirements are input from the project party, design team, or users. These requirements include the usage functions of building spaces, load-bearing capacity, energy efficiency requirements, environmental friendliness, daylighting requirements, fire resistance performance, etc. Design goals are detailedly expressed through functional requirements (room functions, window area, number of stairs), performance requirements (structural strength, seismic requirements, thermal comfort), and regulatory requirements (building safety standards, energy efficiency standards).

[0046] Extract the geometric properties, physical properties, and their spatial positions of all functional components from the BIM model. The input information of functional components will be matched with the building design standard library to ensure that the functional components comply with relevant design specifications (Building Design Code, energy efficiency standards).

[0047] S502, Component-level adjustment based on design rules and reverse algorithm: Preset a design rule library, which contains common building design rules (space layout, mechanical properties, energy efficiency requirements) and adjustment specifications for various building components (dimensions, positions, material selection). The design rule library also includes optimization criteria based on experience or cases to help designers optimize and adjust components according to project characteristics.

[0048] Use a goal-driven reverse algorithm to reverse-derive the adjustment parameters of each component according to the design requirements input by the user. These parameters can include: geometric adjustment (wall thickness, door and window sizes, stair slopes), physical property adjustment (material selection, load-bearing capacity, fire resistance performance), and spatial adjustment (such as room layout, door and window positions, daylighting and ventilation requirements). The reverse algorithm optimally configures building components by establishing an optimization function and combining a multi-objective optimization strategy (particle swarm algorithm). The optimization goals include: ensuring structural safety and stability; improving the energy efficiency, comfort, and environmental adaptability of the building; meeting the functional requirements of the building (space utilization rate, daylighting rate).

[0049] S503, Generation and correction of physical performance requirements: Combine all functional requirements and design rules to generate the physical performance requirements of the BIM model that meet the actual design goals. Check the physical performance of the bone structure (especially load-bearing components) to ensure that it can meet the overall building physical performance requirements. These physical performance requirements include: load-bearing capacity requirements to ensure that the load capacity of each component in the structure meets the expectations; seismic and wind resistance performance, and design the structure accordingly according to the natural environmental conditions (such as earthquakes, winds, etc.) of the building location; thermal engineering and energy efficiency requirements to ensure that the thermal engineering performance (such as heat insulation, ventilation, etc.) of the building meets the requirements according to energy efficiency standards and design specifications; noise control and comfort, and ensure that the building can provide a comfortable noise environment during use by adjusting the sound insulation design.

[0050] Based on the above physical performance requirements, adjust the geometry and material properties of the bone structure and other related components to ensure that the model meets the comprehensive physical performance requirements of the building. During the adjustment process, conduct verification to ensure that each adjustment can enhance the feasibility of the model without violating the basic principles of architectural design.

[0051] S504, Physical Performance Analysis and Verification of the Bone Structure: After adjustment, conduct physical performance analysis on the bone structure, including: Bearing Capacity Analysis: Through structural mechanics simulation, evaluate the response of the adjusted bone components under load to ensure no overloading or damage occurs; Node and Connection Performance: Check the stability of each connection node of the bone structure to confirm that the connection points can transfer loads safely and effectively; Overall Stability: Conduct an overall stability assessment of the entire bone structure to ensure that its earthquake resistance, wind resistance, etc. meet the design standards; Energy Efficiency and Thermal Comfort Analysis: Conduct thermal engineering and energy efficiency analysis on the bone structure and its combined parts with other components to ensure that the building structure meets the energy conservation and emission reduction goals.

[0052] After completing the physical performance analysis, confirm that the final BIM model meets all design requirements and perform corresponding updates and archiving. This version of the BIM model will serve as the basic data for building construction and operation and maintenance.

[0053] Step S6: Perform comprehensive performance analysis on the adjusted BIM model, where the performance analysis includes structural mechanics performance, energy consumption simulation, and natural lighting; match optimization suggestions and corresponding design requirements according to the comprehensive performance analysis results to drive the automatic iterative update of the BIM model; S601, Construction of the Comprehensive Performance Analysis Model: Construct a comprehensive performance analysis framework to convert the various performance requirements of the building (structure, energy efficiency, lighting) into quantifiable analysis parameters.

[0054] The framework includes: Structural Mechanics Performance - Evaluate the stability and bearing capacity of the building structure; Energy Efficiency Simulation - Analyze the energy efficiency, thermal performance, and energy consumption performance of the building; Natural Lighting Analysis - Evaluate the natural lighting situation inside the building to ensure that it meets the comfort and energy conservation goals.

[0055] S602, Structural Mechanics Performance Analysis: Use structural mechanics analysis tools (such as SAP2000) to conduct a comprehensive structural mechanics performance analysis on the adjusted BIM model. The main analysis contents include: Bearing Capacity: Analyze the bearing capacity of each component under different loads to ensure that problems such as yielding and fracture do not occur; Seismic performance: Considering the seismic intensity of the area where the building is located, analyze the response of the building under seismic loads to ensure that the building meets the seismic design requirements; Joint connection and mechanical stability: Through the analysis of connection joints, evaluate the connection methods between building components and the stability of joints to ensure the overall structural stability; Local effects: Conduct local stress analysis on large-span components and special components (such as the facade structure of high-rise buildings, etc.) to ensure that they do not become unstable under extreme load conditions.

[0056] The structural analysis results will provide information such as stress, strain, and displacement, helping to analyze whether there are potential structural safety hazards and providing optimization suggestions based on these results. According to the analysis results, identify the components that need to be adjusted and carry out dimension adjustment, material selection, or strengthening design to improve the safety and stability of the structure.

[0057] S603, Energy efficiency simulation and optimization: Use EnergyPlus to conduct energy efficiency simulation on the building, and analyze the thermal comfort, energy consumption, heating and cooling loads, and air conditioning requirements of the building. The specific analysis contents include: Heat transfer analysis: Calculate the heat transfer performance of building exterior walls, roofs, windows, etc., and analyze the heat insulation effect of the building; Energy efficiency assessment: Based on the structural layout, daylighting, ventilation, insulation effect, etc. of the building, evaluate the overall energy efficiency level, and evaluate the heating and cooling loads of the building according to different seasons and climate conditions; Renewable energy utilization: Analyze the utilization of renewable energy in the building (such as solar energy, wind energy, geothermal energy, etc.) and evaluate its contribution to the building's energy efficiency.

[0058] Based on the energy efficiency simulation results and combined with energy-saving design specifications, provide optimization suggestions. For example, it is necessary to adjust the thermal insulation layer of the building envelope, optimize the orientation of windows, and add solar water heaters.

[0059] S604, Natural daylighting analysis and optimization: Analyze the natural daylighting inside the building through a daylighting simulation tool (Dialux) to ensure that the building's lighting system can make the most of natural light. The main analysis contents include: Sunshine duration and intensity: Analyze the natural daylighting conditions in different areas of the building at different times, especially important functional areas (such as office areas, living rooms); Daylighting ratio and uniformity: Evaluate the uniformity of indoor lighting to avoid problems such as uneven lighting or dark areas; Combination of daylighting and energy efficiency: Combine daylighting analysis with energy efficiency simulation to evaluate the contribution of daylighting optimization to building energy conservation.

[0060] Based on the daylighting analysis results, adjust the window size, orientation, shading measures, etc. of the building to improve daylighting efficiency and spatial comfort. For areas that require enhanced daylighting, consider adding designs such as skylights and solar tubes; for areas with excessive daylighting, adjustments need to be made through shading facilities or curtain designs to reduce energy consumption and improve comfort.

[0061] S605, Comprehensive Performance Evaluation and Feedback Mechanism: Integrate the results of structural mechanics performance, energy efficiency, and daylighting analysis into a comprehensive performance evaluation report. This report will provide the project team with a comprehensive view of performance evaluation, helping decision-makers understand the performance of the building in various aspects. The evaluation results are presented in forms such as charts, heat maps, and simulation videos for designers and clients to intuitively understand the performance of the building.

[0062] Generate optimization suggestions based on the comprehensive performance analysis results, including specific suggestions for structural optimization, energy efficiency improvement, and daylighting enhancement.

[0063] S606, Automatic Iterative Update and Verification of BIM Model: Based on the comprehensive performance analysis results and optimization suggestions, the system will automatically adjust each component in the BIM model, modify geometric and physical properties, and update the adjusted BIM model to the latest version. The update process will include automated component replacement, size adjustment, and material selection to ensure that the BIM model meets the optimized design requirements. The updated BIM model will undergo another structural mechanics analysis, energy efficiency simulation, and daylighting analysis to ensure that the adjusted model meets performance requirements in all aspects. The verification process will ensure that all design requirements of the building are met, while also meeting building codes and project requirements.

[0064] Step S7: Support users to dynamically adjust design requirements, display each BIM model obtained during the optimization process and its comprehensive performance analysis results through a visual interface, achieve rapid "what-if" analysis and scheme comparison, and help users make decisions among multiple design schemes.

[0065] Visualize the results of the comprehensive performance analysis and display the optimized building design in forms such as charts, heat maps, and 3D simulations, enabling the design team to intuitively understand the performance of the building in various aspects. Use the visual interface in the BIM platform to display the effects after model update, helping designers and clients view the advantages and disadvantages of different design schemes and quickly compare the schemes.

[0066] Supports users for dynamic adjustment. Through "simulation hypothesis" analysis, users can quickly compare and select among multiple different design schemes, view the impact of the adjustment on building performance, and help make the optimal decision among multiple schemes. According to different optimization objectives and design requirements, the system will provide multiple design schemes for decision-makers to choose from and evaluate the comprehensive performance of each scheme in real time.

[0067] Embodiment 2, a three-dimensional digital survey and design system based on BIM, see Figure 1 as shown, including the following modules: Data acquisition module, which acquires the point cloud data of the target building and the corresponding material information, and generates an initial three-dimensional spatial data model; Component recognition module, based on the three-dimensional spatial data model, uses a component recognition model to automatically identify and classify the main components in the building, extracts the geometric and physical properties of the components, and establishes a component attribute database; Physical performance analysis module, which conducts physical performance analysis on the bone components in combination with the material information, and splices the bone components into a bone structure model according to the analysis results; BIM modeling module, which registers and replaces the data of the bone structure model and functional components in the three-dimensional spatial data model, calls the preset BIM component family template, automatically generates a BIM model, and obtains the bone framework; Reverse adjustment module, based on the preset design rules and reverse deduction algorithm, conducts component-level adjustment on the existing BIM model, generates the physical performance requirements of the BIM model, and analyzes the bone framework to ensure that it meets the physical performance requirements of the BIM model; Performance analysis and optimization module, which conducts comprehensive performance analysis on the adjusted BIM model, matches optimization suggestions according to the analysis results, and drives the automatic iterative update of the BIM model; User interaction module, which supports users to dynamically adjust the design requirements, and displays each BIM model and its comprehensive performance analysis results during the optimization process through a visual interface.

[0068] The specific embodiments described above further elaborate on the purpose, technical solutions, and beneficial effects of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not used to limit the protection scope of the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope 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 processing to generate an initial three-dimensional 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 properties and physical properties of the components are extracted to establish a component property database; the classification includes skeleton components and functional components; The member retains two marking points for registration; Performing physical property analysis on the bone components in combination with corresponding material information, and splicing the bone components into a bone structure model according to the physical property analysis results; Register and replace the bone structure model and functional components in the three-dimensional space data model; Call the preset BIM component family template, automatically construct the 3D spatial data model after data replacement into a BIM model, and mark the architecture corresponding to the bone structure model in the BIM model as the bone architecture of the BIM model; The process of automatically constructing the BIM model complies with architectural design standards and specifications; The design requirements are obtained by combining functional components. Based on the preset design rules and reverse algorithm, the existing BIM model is adjusted at the component level, and the physical performance requirements of the BIM model are generated. After the adjustment, the physical performance analysis of the skeleton is performed to ensure that it meets the physical performance requirements of the BIM model.

2. A three-dimensional digital survey and design method based on BIM 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 BIM models; 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 three-dimensional digital survey and design method based on BIM according to claim 1 is characterized in that: The structure of the component identification model includes: The input layer is used to obtain the three-dimensional spatial data model of the building; The feature extraction layer is used to extract the multi-scale spatial features of the 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 attributes of each component and obtain the physical attributes 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, using a staged classification approach to first identify all bone components and then functional components; The adjacency association analysis layer is used to analyze the spatial adjacency relationship and structural connection relationship 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 three-dimensional digital survey and design method based on BIM according to claim 3, characterized in that: The training of the component recognition model includes: The three-dimensional spatial data model of the building with complete annotated bone components 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 the few-sample learning method combined with the initial training set to obtain an initial component recognition model; a self-training mechanism is used to combine the three-dimensional spatial data models of a large number of unlabeled buildings for incremental training, and 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 components and bone components; the training results are evaluated for accuracy through cross-validation and structural adjacent relationship analysis, and finally the optimal model parameter set for component recognition is determined to obtain the final component recognition model.

5. The three-dimensional digital survey and design method based on BIM according to claim 1 is characterized in that: 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 bearing capacity, bending stiffness, node connection performance and material strength index. According to the analysis results, bone component pairs that meet the structural safety and connection feasibility are screened, 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 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 relationship that does not meet the performance requirements is locally optimized and adjusted until a bone structure model that meets the target physical performance requirements is generated.

6. The BIM-based three-dimensional digital survey and design method according to claim 1, characterized in that: 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 a preset BIM component family template, wherein 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 building 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.

7. The three-dimensional digital survey and design method based on BIM according to claim 1, characterized in that: The component-level adjustment includes: The design goal input by the user is obtained, the goal-driven rules and component constraint rules in the design rule library are called, and the optimization function is established in combination with the multi-objective optimization algorithm. The component-level change suggestions that meet the design goals are generated 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.

8. 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 7, including: 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, based on the three-dimensional spatial data model, uses the component recognition model to automatically identify and classify the main components in the building, extracts the geometric and physical properties of the components, and establishes a component property database; The physical performance analysis module analyzes the physical performance of the bone components in combination with the material information, and splices the bone components into a bone structure model according to the analysis results; The BIM modeling module aligns and replaces the bone structure model and functional components in the three-dimensional space 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 the preset design rules and reverse algorithm, 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 the automatic iterative update 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 in the optimization process through a visual interface.

Images