An artificial intelligence-based automatic generation method and system for architectural design scenes

Abstract

The present application provides a method and system for automatically generating architectural design scenes based on artificial intelligence. Specifically, by integrating the geometric feature parameter library of special-shaped components and the spatial topological relationship parameter library, an artificial intelligence optimization engine is used for joint analysis to create a dynamic morphological constraint map of the architectural scene. The process includes using laser point cloud scanning to obtain three-dimensional data of the construction site, and extracting real-time curvature gradient parameters and functional area connection node parameters from it. These parameters are then matched and verified with the corresponding thresholds and conditions in the dynamic morphological constraint map to generate an adaptive correction parameter set for the architectural scene. Finally, based on this parameter set, the control points of the processing trajectory of the building components are adjusted to form construction instructions that include the linkage between spatial morphology and functional topology. The technical solution provided by the present application can improve the efficiency and flexibility of the automatic generation of architectural design scenes.

Description

Technical Field

[0001] The present application relates to the technical field of automatic generation of architectural design scenes, and in particular to an artificial intelligence-based automatic generation method and system for architectural design scenes. Background Art

[0002] In modern architectural design and construction, the precise handling of complex curved surfaces and multifunctional spaces places higher technical demands. As architectural forms become increasingly diverse and complex, designers must not only consider structural aesthetics but also ensure a rational functional layout, efficient space utilization, and adaptability to ever-changing construction conditions. This is particularly true in the design and fabrication of special-shaped building components. There is an urgent need for a technology that enables real-time spatial morphology analysis and functional topology optimization to support automated management of the entire design and construction process.

[0003] Parametric modeling combined with finite element analysis (FEA) is now widely used in modern architectural design and construction. This approach allows designers to generate complex geometries by defining a series of control parameters and evaluate the mechanical properties of the structure using FEA tools. Furthermore, it supports integrating design models with the actual construction environment, thus addressing the issue of translating design intent into construction instructions. This approach leverages advanced computing technology and engineering expertise, making the design process more scientific and rational while also improving the quality and safety of the final product.

[0004] However, while the above solution meets the needs of architectural design to a certain extent, it still has some significant flaws. First, because it relies on pre-set parameter sets and assumptions, it is difficult to cope with the uncertainty brought about by dynamic changes on the construction site. Second, finite element analysis is usually time-consuming, which is not conducive to rapid iterative design and immediate adjustment of construction plans. Finally, when dealing with complex surfaces and changing functional area connection points, this method often requires a lot of manual intervention to adjust and optimize, which limits the improvement of overall work efficiency. These shortcomings highlight the shortcomings of existing technical solutions in terms of flexibility and real-time performance. Summary of the Invention

[0005] The present application provides an artificial intelligence-based method and system for automatically generating architectural design scenes, which is used to solve the problems of low efficiency and poor flexibility in the existing technology for automatically generating architectural design scenes.

[0006] In a first aspect, the present application provides a method for automatically generating architectural design scenarios based on artificial intelligence, comprising:

[0007] Obtaining a geometric feature parameter library of special-shaped components including a surface curvature gradient change threshold and a spatial topological relationship parameter library including functional area boundary constraint conditions for architectural design scenarios;

[0008] The artificial intelligence optimization engine is used to jointly analyze the surface curvature gradient change threshold and the functional area boundary constraint conditions to generate a dynamic morphological constraint map of the building scene;

[0009] Acquire three-dimensional space dot matrix data of the construction site in real time through laser point cloud scanning, and extract real-time curvature gradient parameters and functional area connection node parameters from the three-dimensional space dot matrix data;

[0010] Performing spatial morphological matching between the real-time curvature gradient parameter and the surface curvature gradient change threshold in the dynamic morphological constraint map of the building scene, and performing topological relationship verification between the functional area connection node parameter and the functional area boundary constraint condition, to generate a building scene adaptive correction parameter set;

[0011] The processing trajectory control points of the building components are reconstructed based on the building scene adaptive correction parameter set, and the building design scene construction instructions containing the linkage of spatial form and functional topology are generated.

[0012] Optionally, the artificial intelligence optimization engine is used to jointly analyze the surface curvature gradient change threshold and the functional area boundary constraint condition to generate a dynamic morphological constraint map of the building scene, including:

[0013] Decomposing the surface curvature gradient change threshold into a curvature gradient feature vector, and decoupling the functional area boundary constraint condition into a multi-dimensional topological constraint feature vector;

[0014] Mapping the curvature gradient eigenvector and the multidimensional topological constraint eigenvector to the same coupling analysis space, and generating a hybrid constraint tensor that links geometric form and spatial topology through tensor fusion;

[0015] Constructing a dynamic constraint diffusion model based on the hybrid constraint tensor, driving the node propagation path of the multidimensional topological constraint feature vector with the spatial attenuation characteristics of the curvature gradient feature vector to form a constraint relationship diffusion network;

[0016] The constraint relationship diffusion network is reversely corrected and its boundaries are extracted. A hierarchical dynamic morphological constraint boundary is generated according to the spatial continuity of the curvature gradient sensitive domain and the functional topological conduction chain, and then reorganized into a dynamic morphological constraint map of the architectural scene.

[0017] Optionally, the step of driving the node propagation path of the multidimensional topological constraint feature vector using the spatial attenuation characteristic of the curvature gradient feature vector to form a constraint relationship diffusion network includes:

[0018] Converting the curvature gradient eigenvector in the hybrid constraint tensor into a spatial attenuation weight matrix, and converting the multidimensional topological constraint eigenvector into a topological connection probability field;

[0019] Dynamically weighting the topological connection probability field based on the spatial attenuation weight matrix to generate a directional constrained diffusion channel, wherein the activation intensity of the constrained diffusion channel is dynamically adjusted by the relationship between the spatial attenuation coefficient and the topological connection probability;

[0020] Adjusting the path connectivity priority of the topological connection probability field according to the distribution gradient of the spatial attenuation weight to generate a constrained diffusion network extending from a high attenuation area to a low attenuation area;

[0021] The constrained diffusion network is hierarchically clustered based on the constrained diffusion channel, and cross-hierarchical conduction links are established through overlapping nodes to form a constrained relationship diffusion network.

[0022] Optionally, adjusting the path connectivity priority of the topological connection probability field according to the distribution gradient of the spatial attenuation weight to generate a constrained diffusion network extending from a high attenuation area to a low attenuation area includes:

[0023] The spatial attenuation weight is mapped to a spatial gradient field, where the gradient amplitude of each node is the norm of the weight difference of the adjacent nodes, and the gradient direction is the neighborhood direction where the weight decreases fastest;

[0024] Constructing a directed connected graph based on the gradient direction, retaining for each node only the adjacent gradient edges pointing to nodes with a gradient magnitude lower than the current node, and filtering the reverse gradient edges;

[0025] According to the initial connection probability of each node in the topological connection probability field, the normalized inverse of the gradient amplitude is superimposed as a path selection weight to generate a decay-aware connection priority score;

[0026] The nodes are traversed from high to low according to the connection priority scores, and starting from the seed node with the maximum spatial attenuation weight, the connection edges are dynamically expanded along the gradient descent direction to generate a constrained diffusion network extending from the high attenuation area to the low attenuation area.

[0027] Optionally, the performing of topological relationship verification on the functional area connection node parameters and the functional area boundary constraint conditions to generate a building scene adaptive correction parameter set includes:

[0028] The real-time curvature gradient parameters are parameterized in manifold space, an orthogonal parameter coordinate system is established along the main curvature direction of the surface, and the continuous gradient changes are converted into discretized morphological deviation tensors;

[0029] A dynamic constraint network of the functional area connection node parameters is established through a spatial topological relationship parameter library, and a surface morphology correction field is generated according to the morphological deviation tensor and the functional area boundary constraint conditions;

[0030] Injecting the functional area connection node parameters into the dynamic constraint network, performing topological energy diffusion calculation based on the constraint propagation path of the functional area connection node, and generating a topological correction gradient distribution cloud map;

[0031] A joint optimization space of the surface morphology correction field and the topology correction gradient distribution cloud map is established, and a potential energy surface of the joint action of morphology and topology is constructed through dual-parameter coupling. The parameter space is corrected along the descending direction of the potential energy gradient to generate an adaptive correction parameter set for the building scene.

[0032] Optionally, establishing a joint optimization space of the surface morphology correction field and the topology correction gradient distribution cloud map, and constructing a potential energy surface of joint morphology and topology through dual-parameter coupling, includes:

[0033] The surface morphology correction field is spatially discretized to generate a gridded morphology correction parameter matrix covering the building scene surface, and the topology correction gradient distribution cloud map is converted into a node gradient intensity tensor of the same spatial resolution;

[0034] Establishing a spatial mapping relationship between the gridded morphology correction parameter matrix and the node gradient intensity tensor, and forming a joint parameter space through parameter coordinate registration;

[0035] Defining a two-variable coupling relationship between the surface morphology correction field and the topology correction gradient distribution cloud in the joint parameter space, and constructing a two-parameter coupling relationship reflecting the interaction between morphological deformation energy and topological constraint energy;

[0036] The two-parameter coupling relationship is fully differentially expanded, and the surface curvature continuity and topological constraint propagation characteristics are combined to construct the potential energy surface of the joint action of morphology and topology.

[0037] Optionally, reconstructing the processing trajectory control points of the building components based on the building scene adaptive correction parameter set to generate the building design scene construction instructions including the linkage between spatial form and functional topology includes:

[0038] Decomposing the building scene adaptive correction parameter set into a morphological correction parameter matrix and a topological correction gradient tensor, wherein the morphological correction parameter matrix includes three-dimensional displacement components of surface control points;

[0039] Performing spatiotemporal discretization processing on the morphological correction parameter matrix based on a parameterized coordinate system, and generating an initial distribution of control points of the machining trajectory according to bicubic spline interpolation;

[0040] Mapping the topology correction gradient tensor to the space where the control points of the machining trajectory are initially distributed, applying topology correction weights to the control points along the gradient association channel to form a composite control point field;

[0041] A trajectory optimization model is established in the composite control point field, trajectory mutations between control points are eliminated through local curvature adaptive adjustment, and topology correction gradients are transferred to adjacent control point clusters to construct a hierarchical control network;

[0042] The control point clusters of the key feature areas are spatially coupled with the topology correction gradient to generate construction instructions for architectural design scenarios that include the linkage between spatial form and functional topology.

[0043] In a second aspect, the present application provides an artificial intelligence-based automatic generation system for architectural design scenarios, comprising:

[0044] An acquisition module is used to obtain a geometric feature parameter library of special-shaped components including a surface curvature gradient change threshold and a spatial topological relationship parameter library including functional area boundary constraint conditions of the architectural design scene;

[0045] An analysis module, which uses an artificial intelligence optimization engine to jointly analyze the surface curvature gradient change threshold and the functional area boundary constraint conditions to generate a dynamic morphological constraint map of the building scene;

[0046] An extraction module acquires three-dimensional space dot matrix data of the construction site in real time through laser point cloud scanning, and extracts real-time curvature gradient parameters and functional area connection node parameters from the three-dimensional space dot matrix data;

[0047] A verification module performs spatial morphological matching on the real-time curvature gradient parameter and the surface curvature gradient change threshold in the dynamic morphological constraint map of the building scene, and simultaneously verifies the topological relationship between the functional area connection node parameter and the functional area boundary constraint condition, thereby generating a building scene adaptive correction parameter set;

[0048] A generation module reconstructs the processing trajectory control points of the building components based on the building scene adaptive correction parameter set, and generates construction instructions for the building design scene that include the linkage between spatial form and functional topology.

[0049] In a third aspect, the present application provides a computing device comprising a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement an artificial intelligence-based automatic generation method of architectural design scenes as described in the first aspect above.

[0050] In a fourth aspect, the present application provides a computer storage medium storing a computer program. When the computer program is executed by a computer, it implements an artificial intelligence-based automatic generation method for architectural design scenes as described in the first aspect.

[0051] In an embodiment of the present application, a geometric feature parameter library of special-shaped components including a surface curvature gradient change threshold and a spatial topological relationship parameter library including functional area boundary constraints of an architectural design scene are obtained; the surface curvature gradient change threshold and the functional area boundary constraints are jointly analyzed by an artificial intelligence optimization engine to generate a dynamic morphological constraint map of the architectural scene; three-dimensional spatial lattice data of the construction site is acquired in real time by laser point cloud scanning, and real-time curvature gradient parameters and functional area connection node parameters in the three-dimensional spatial lattice data are extracted; the real-time curvature gradient parameters are spatially matched with the surface curvature gradient change threshold in the dynamic morphological constraint map of the architectural scene, and the topological relationship between the functional area connection node parameters and the functional area boundary constraints is verified to generate an adaptive correction parameter set for the architectural scene; the processing trajectory control points of the architectural components are reconstructed based on the adaptive correction parameter set for the architectural scene to generate construction instructions for the architectural design scene that include the linkage between spatial morphology and functional topology.

[0052] The technical solution of this application has the following beneficial effects:

[0053] This application establishes a parameter library for the geometric characteristics of irregularly shaped components, including thresholds for surface curvature gradient changes, and a parameter library for spatial topological relationships for functional area boundary constraints. This library accurately describes complex shapes and spatial layouts in architectural design, providing fundamental data support for subsequent optimization. By jointly analyzing data from these two libraries using an artificial intelligence optimization engine, a dynamic morphological constraint map is intelligently generated that reflects the balance between design requirements and physical constraints, enhancing the rationality and innovation of architectural design solutions. Laser point cloud scanning technology is used to extract three-dimensional spatial dot matrix data from the construction site in real time, from which real-time curvature gradient parameters and functional area connection node parameters are calculated. This ensures that the design solution can be promptly adapted to the actual conditions of the construction site and improves the efficiency of the design-to-construction transition. By matching and verifying on-site data with a pre-generated dynamic morphological constraint map, a modified parameter set adapted to the actual construction environment is automatically generated, ensuring the accuracy and feasibility of the architectural design during implementation. This ultimately generates detailed instructions to guide construction, enabling an integrated process from design to manufacturing to construction, significantly improving the overall quality and efficiency of the construction project.

[0054] Furthermore, the generation process of the dynamic morphological constraint map for architectural scenes was refined. Specifically, the curvature gradient change threshold and functional area boundary constraints were deconstructed into feature vectors, and a hybrid constraint tensor was created within a unified space using tensor fusion techniques to construct a dynamic constraint diffusion model. This model drives node propagation paths based on spatial attenuation characteristics, forming hierarchical dynamic morphological constraint boundaries that reflect spatial continuity. This process effectively combines the requirements of geometric form and spatial topology, making the generated dynamic morphological constraint map for architectural scenes more refined and accurate, greatly enhancing the flexibility and adaptability of architectural design while improving the quality and practicality of design results.

[0055] These and other aspects of the present application will become more readily apparent from the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0057] Figure 1 A flowchart of an artificial intelligence-based automatic generation method for architectural design scenes provided by the present application is shown;

[0058] Figure 2 The present invention provides a schematic diagram of the structure of an artificial intelligence-based automatic generation system for architectural design scenes;

[0059] Figure 3 A schematic structural diagram of a computing device provided by the present application is shown. DETAILED DESCRIPTION

[0060] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0061] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this document or may be executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any order of execution. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this document are used to distinguish different messages, devices, modules, etc., and do not represent a sequential order, nor do they limit "first" and "second" to being different types.

[0062] This solution integrates a library of geometric feature parameters for special-shaped components and a library of spatial topological relationship parameters. Based on an artificial intelligence optimization engine, it jointly analyzes the threshold of surface curvature gradient changes and the boundary constraints of functional areas in architectural design scenarios, thereby generating a dynamic morphological constraint map for architectural scenarios and realizing the automation and intelligence of architectural design.

[0063] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0064] Figure 1 The present invention provides a flowchart of a method for automatically generating architectural design scenes based on artificial intelligence. Figure 1 As shown, the method includes:

[0065] 101. Obtain a geometric feature parameter library of special-shaped components including a threshold value for a surface curvature gradient change and a spatial topological relationship parameter library including boundary constraints for functional areas in the architectural design scene;

[0066] In this step, the surface curvature gradient change threshold refers to the maximum curvature change range allowed on a specific surface in architectural design. It is used to control the change speed and smoothness of the surface shape to ensure that the building appearance is both beautiful and structurally stable.

[0067] The geometric parameter library for irregularly shaped components is a data set that stores detailed information about irregularly shaped parts in buildings. This information includes the specific dimensions and shapes of each component, as well as variations in curvature, such as maximum and minimum curvatures, and the rate of curvature change. This data guides the precise modeling of complex shapes during design and construction.

[0068] The spatial topological relationship parameter library of functional area boundary constraints records the spatial connection methods and boundary restrictions between different functional areas in the building, such as the positional relationship and channel width of functional areas such as meeting rooms and rest areas, to ensure that the layout is reasonable and meets usage requirements.

[0069] In the embodiment of the present application, first, all special-shaped components in the architectural design are scanned with high precision using 3D scanning technology to collect their detailed geometric information, especially the curvature gradient change threshold. Next, these raw data are converted into digital models using professional software and classified and sorted to form a library of geometric feature parameters of special-shaped components. Then, for the boundary constraints of functional areas, the architectural design drawings or models are analyzed to extract the position coordinates, size, relative position relationship between the two, and necessary boundary conditions of all functional areas. Finally, the above two types of data are integrated into their respective corresponding databases to provide accurate basic data support for subsequent design optimization.

[0070] 102. Using an artificial intelligence optimization engine, jointly analyze the surface curvature gradient change threshold and the functional area boundary constraint conditions to generate a dynamic morphological constraint map of the building scene;

[0071] In this step, the AI ​​optimization engine is a technical system that can process large amounts of data and find the best solution.

[0072] The dynamic morphological constraint graph for architectural scenes is a set of rules generated by the engine that describe the specific morphological and spatial relationships that must be adhered to during architectural design. This set of rules takes into account the threshold of surface curvature gradient changes and the boundary constraints of functional areas, aiming to ensure that architectural designs are both aesthetically pleasing and practical.

[0073] In this embodiment of the present application, the two parameter libraries created in step 101 are first imported into the artificial intelligence optimization engine. Next, a deep learning algorithm is used to jointly analyze the threshold for surface curvature gradient changes and the functional area boundary constraints to identify potential design conflicts and optimization opportunities. Based on the analysis results, a series of constraints describing the building form and spatial relationships are then created, such as the maximum permissible curvature range of the surface and the minimum spacing between functional areas. Finally, these constraints are combined into a comprehensive dynamic morphological constraint map, which serves as an important reference for subsequent design stages.

[0074] 103. Acquire three-dimensional space dot matrix data of the construction site in real time through laser point cloud scanning, and extract real-time curvature gradient parameters and functional area connection node parameters from the three-dimensional space dot matrix data;

[0075] In this step, laser point cloud scanning is a technology that uses a laser beam to measure the surface position of an object. The generated three-dimensional space dot matrix data contains the position coordinate information of each point in the construction site.

[0076] The real-time curvature gradient parameter is a key indicator of the change in the curvature of the site surface extracted from these coordinate data and is used to evaluate the changes in the terrain.

[0077] Functional area connection node parameters are data that identify the specific connection locations between different functional areas, such as the specific coordinates and size information of channels, doors, etc., to ensure that the connectivity between functional areas meets the design requirements.

[0078] In the embodiment of the present application, first, a laser scanning device is deployed at the construction site for a comprehensive scan to collect three-dimensional spatial dot matrix data of the construction site. Next, a data analysis tool is used to extract real-time curvature gradient parameters and functional area connection node parameters from these data. These parameters are then compared and analyzed with the corresponding data in the parameter library established in the early design phase to verify whether the actual situation on site is consistent with the design. Finally, the construction plan or design parameters are adjusted based on the comparison results to ensure that the actual construction can accurately reflect the requirements of the design scheme.

[0079] 104. Perform spatial morphological matching between the real-time curvature gradient parameter and the surface curvature gradient change threshold in the dynamic morphological constraint map of the building scene, and perform topological relationship verification between the functional area connection node parameter and the functional area boundary constraint condition to generate a building scene adaptive correction parameter set;

[0080] In this step, the architectural scene dynamic form constraint map is a set of rules describing the architectural form and spatial relationship generated based on the data in the early stage of architectural design.

[0081] Spatial morphology matching refers to comparing the real-time curvature gradient parameters collected on-site with the surface curvature gradient change threshold in the constraint map to verify whether the on-site conditions meet the design requirements.

[0082] Topological relationship verification is to check whether the parameters of the functional area connection nodes meet the boundary constraints specified in the design, ensuring that the layout of each functional area is reasonable and the connectivity is good.

[0083] In the embodiment of the present application, first, the real-time curvature gradient parameters obtained in step 103 are compared with the surface curvature gradient change threshold in the building scene dynamic morphological constraint map to identify any inconsistencies. Next, necessary corrections or design adjustments are made to the problems found. Then, the functional area connection node parameters are verified against the functional area boundary constraints to ensure that the position and size of all connection points meet the design standards. Finally, the above analysis results are integrated to generate a set of building scene adaptive correction parameters to provide guidance for subsequent construction.

[0084] 105. Reconstruct the processing trajectory control points of the building components based on the building scene adaptive correction parameter set to generate building design scene construction instructions that include spatial form and functional topology linkage.

[0085] In this step, the building scene adaptive correction parameter set is a set of parameters that are adjusted from the original design scheme according to the actual situation on site, including the morphological correction parameter matrix and the topological correction gradient tensor.

[0086] Processing trajectory control points are specific coordinate points used to guide the movement path of robotic arms or other automated equipment during the manufacturing process of building components.

[0087] The construction instructions for architectural design scenarios that incorporate the linkage between spatial form and functional topology refer to the final construction guidelines adjusted based on the actual on-site conditions, ensuring that the spatial form and functional layout of the building meet both aesthetic requirements and practical usage needs.

[0088] In an embodiment of the present application, first, the architectural scene adaptive correction parameter set is decomposed into a morphological correction parameter matrix and a topological correction gradient tensor, and the initial distribution of the processing trajectory control points is generated according to the bicubic spline interpolation method. Next, the topological correction gradient tensor is mapped to the space of the processing trajectory control points, and correction weights are applied to the control points along the gradient association channel to form a composite control point field. Then, a trajectory optimization model is established in the composite control point field, and the trajectory mutations between the control points are eliminated by adaptively adjusting the local curvature, and the topological correction gradient is transferred to the adjacent control point clusters to construct a hierarchical control network. Finally, the control point clusters and the topological correction gradients in the key feature areas are combined to generate architectural design scene construction instructions that include the linkage between spatial morphology and functional topology.

[0089] In summary, steps 101 to 105, from preliminary architectural design planning to construction guidance, seamlessly transitioned the design concept to the actual built product. This not only enhanced the scientific and rational nature of the architectural design but also significantly improved construction efficiency and quality, ensuring that the building both embodies aesthetic value and meets the diverse needs of practical use. Each step was closely linked, contributing to the successful implementation of the project and ultimately delivering a building that is both beautiful and practical.

[0090] In step 102, the artificial intelligence optimization engine is used to jointly analyze the surface curvature gradient change threshold and the functional area boundary constraint condition to generate a dynamic morphological constraint map of the building scene, including:

[0091] 201. Decomposing the surface curvature gradient change threshold into a curvature gradient feature vector, and decoupling the functional area boundary constraint condition into a multi-dimensional topological constraint feature vector;

[0092] In step 201, the curvature gradient feature vector is a dataset derived from decomposing the surface curvature gradient change threshold. It contains information describing the degree of curvature of the building surface and its rate of change. The multidimensional topological constraint feature vector is a dataset decoupled from the functional area boundary constraints, describing the spatial connectivity and constraints between different functional areas. These parameters are used for precise adjustments during the subsequent design process to ensure that the design meets physical and functional requirements.

[0093] In the present embodiment, a mathematical algorithm is first used to convert the surface curvature gradient change threshold into a series of numerically represented curvature gradient feature vectors. Next, graphics processing techniques are used to decouple the functional area boundary constraints into multidimensional topological constraint feature vectors. Data analysis methods are then used to ensure that each feature vector accurately reflects the characteristics of its corresponding design element. Finally, the two sets of feature vectors are combined to prepare for the next step of fusion analysis.

[0094] 202. Map the curvature gradient eigenvector and the multidimensional topological constraint eigenvector to the same coupling analysis space, and generate a hybrid constraint tensor that links geometric form and spatial topology through tensor fusion;

[0095] In step 202, the coupled analysis space is a virtual spatial environment in which geometric forms and spatial topological relationships can be processed simultaneously. A hybrid constraint tensor is a data structure generated by mapping curvature gradient eigenvectors and multidimensional topological constraint eigenvectors into the same space using tensor fusion technology. It is used to describe complex relationships in architectural design. This tensor integrates both geometric form and spatial topological information, providing basic data support for the subsequent construction of a dynamic constraint diffusion model.

[0096] In this embodiment, the two eigenvectors are first mapped into the same virtual coupled analysis space. Next, a tensor fusion algorithm is applied to generate a hybrid constraint tensor based on the correlation between the eigenvectors. This tensor then integrates both geometric morphology and spatial topology information, providing basic data support for the subsequent construction of a dynamic constraint diffusion model. Ultimately, this process achieves a comprehensive understanding and optimization of architectural design elements.

[0097] 203. Construct a dynamic constraint diffusion model based on the hybrid constraint tensor, and use the spatial attenuation characteristics of the curvature gradient feature vector to drive the node propagation path of the multi-dimensional topological constraint feature vector to form a constraint relationship diffusion network;

[0098] In step 203, the dynamic constraint diffusion model is a simulation system built based on a hybrid constraint tensor, which is used to demonstrate how the curvature gradient eigenvector affects the node propagation path of the multidimensional topological constraint eigenvector, forming a constraint relationship diffusion network. This network reflects the interaction between the various parts of the architectural design. The model helps to identify potential design conflicts and make optimization suggestions. The curvature gradient eigenvector refers to a data set formed by quantifying the information of the change of surface curvature in the architectural design. It contains specific values ​​of the degree of curvature of the building surface and its rate of change, and is used to accurately describe the morphological characteristics of the building facade or interior space. The spatial attenuation characteristic describes how the curvature gradient eigenvector gradually weakens with increasing distance, thereby affecting the design parameters of the surrounding area. The node propagation path of the multidimensional topological constraint eigenvector is a data set generated based on the spatial connection mode and constraint conditions between different functional areas in the architectural design.

[0099] In the present embodiment, a dynamic constraint diffusion model is first established based on a hybrid constraint tensor. Next, the model parameters are set to reflect the effect of the spatial attenuation characteristics of the curvature gradient eigenvector on the multidimensional topological constraint eigenvector. The model is then run to observe changes in node propagation paths and record the results. Finally, the resulting constraint relationship diffusion network provides a basis for subsequent reverse corrections, helping to identify and resolve potential design issues.

[0100] 204. Perform reverse correction and boundary extraction on the constraint relationship diffusion network, generate hierarchical dynamic morphological constraint boundaries based on the spatial continuity of the curvature gradient sensitive domain and the functional topological conduction chain, and reorganize them into a dynamic morphological constraint map of the architectural scene.

[0101] In step 204, the hierarchical dynamic form constraint boundary is obtained by reverse-correcting and extracting the constraint diffusion network, ensuring that the architectural design meets specific aesthetic and technical requirements. The reconstructed architectural scene dynamic form constraint map is a comprehensive set of rules that guide the entire architectural design process. This step ensures that all design elements meet pre-set standards and optimizes the overall layout.

[0102] In this embodiment, the generated constraint diffusion network is first reverse-corrected to identify and adjust any inappropriate components. Next, boundary conditions are extracted to ensure that all design elements meet pre-set standards. Then, hierarchical dynamic morphological constraint boundaries are generated based on the corrected network. Finally, these boundary conditions are reorganized into a final dynamic morphological constraint map for the architectural scene, providing detailed guidance for actual construction.

[0103] In summary, steps 201 to 204, through precise data analysis and intelligent optimization, achieve seamless integration of architectural design from concept to actual operation, greatly enhancing the feasibility and aesthetics of the design scheme. At the same time, it also significantly improves construction quality and efficiency, ensuring that every detail can accurately realize the designer's original intention, making the final building both visually impactful and practical and efficient.

[0104] The step 203 of using the spatial attenuation characteristics of the curvature gradient feature vector to drive the node propagation path of the multi-dimensional topological constraint feature vector to form a constraint relationship diffusion network includes:

[0105] 301. Convert the curvature gradient eigenvector in the hybrid constraint tensor into a spatial attenuation weight matrix, and convert the multidimensional topological constraint eigenvector into a topological connection probability field;

[0106] In step 301, the spatial attenuation weight matrix is ​​a dataset derived from the curvature gradient eigenvectors in the hybrid constraint tensor. It reflects the phenomenon that the curvature of the building surface decreases with distance. The topological connection probability field is a dataset derived from the multidimensional topological constraint eigenvectors. It describes the spatial connection possibilities and constraints between different functional areas. These parameters are used to guide precise adjustments in the subsequent design process to ensure that the design meets physical and functional requirements.

[0107] In the examples of this application, a mathematical algorithm is first used to convert the curvature gradient eigenvector into a spatial attenuation weight matrix to quantify the rate of change of the curvature degree. Next, graphics processing techniques are used to convert the multidimensional topological constraint eigenvector into a topological connection probability field to evaluate the connection probability between different functional areas. Data analysis methods are then used to ensure that each matrix and field accurately reflects the characteristics of its corresponding design element. Finally, these two sets of data are combined to prepare for the next step of dynamic weighted analysis.

[0108] 302. Dynamically weighting the topological connection probability field based on the spatial attenuation weight matrix to generate a directional constrained diffusion channel, wherein the activation intensity of the constrained diffusion channel is dynamically adjusted by the relationship between the spatial attenuation coefficient and the topological connection probability;

[0109] In step 302, dynamic weighting refers to the process of adjusting the topological connection probability field based on the spatial attenuation weight matrix. The constrained diffusion channel (CDC) is a path structure generated by dynamically weighting the topological connection probability field based on the spatial attenuation weight matrix. It describes how the curvature gradient eigenvector in architectural design affects the node propagation path of the multidimensional topological constraint eigenvector. It is directional and reflects the interactions and influence relationships between design elements. The activation intensity of the CDC is dynamically adjusted by the relationship between the spatial attenuation coefficient and the topological connection probability, ensuring that the design meets specific functional and aesthetic requirements.

[0110] Specifically, the spatial attenuation coefficient is used to quantify the attenuation intensity of the building surface curvature gradient as the distance increases, reflecting the rate at which the impact of the curvature on the adjacent area is reduced. The specific calculation formula is:

[0111] ;

[0112] in, refers to the curvature gradient feature vector of the current node (extracted in step 301), It refers to the node To Node The spatial Euclidean distance of Refers to the attenuation rate adjustment factor (preset according to material properties, such as concrete , steel structure ).

[0113] The spatial attenuation weight matrix can be constructed by the above calculation formula .

[0114] Furthermore, the topological connection probability describes the spatial connectivity possibility between different functional areas (such as corridor-hall, equipment room-office area), and is generated by converting the multi-dimensional topological constraint feature vector. The specific calculation steps are:

[0115] The first step is to perform eigendecomposition: multidimensional topological constraint eigenvectors Perform principal component analysis (PCA) to extract key dimensions ;in It is the core functional dimension extracted from the original multidimensional topological constraint feature vector by principal component analysis (PCA). Each dimension represents a class of spatial functional attributes, which can be set according to requirements. For example, : Connectivity strength (such as corridor width, lobby area), : Functional compatibility (such as the affinity between the office area and the conference room), : Crowd density weight (such as hospital emergency area weight , warehouse weight ), : Safety isolation level (such as isolation requirements between laboratories and residential areas).

[0116] The second step is probability field modeling, which is achieved through the following formula:

[0117] ;

[0118] in, is a node To Node The topological connection probability of is the cosine similarity function, with a value range of , for example, similarity > 0.7 → strong functional association (such as ward-nurse station), Representation node The reduced dimension feature vector of ), similarly, Representation node The reduced dimension feature vector of is the probability sensitivity parameter (default ), the larger the value, the more sensitive the probability is to changes in similarity.

[0119] Finally, the activation strength of the constrained diffusion channel is determined based on the spatial attenuation coefficient and topological connection probability determined above. Specifically, the activation strength formula of the constrained diffusion channel can be used:

[0120] ;

[0121] Among them, the adjustment logic is high attenuation area : Forced reduction of activation strength , suppress unnecessary connections; high probability area : Enhance channel activation intensity in low attenuation areas; Indicates the gradient adjustment factor, the default value = 0.5 (calibrated by experiment), the gradient term : Dynamically adjust the weight according to the distribution gradient of the spatial attenuation weight (step 303) to ensure that the path points preferentially to the low attenuation direction.

[0122] In the embodiment of the present application, first, the model parameters are set based on the spatial attenuation weight matrix to reflect its impact on the topological connection probability field. Next, a dynamic weighting algorithm is applied to adjust the path connectivity priority of the topological connection probability field according to the spatial attenuation weight. Then, the model is run to observe the changes in the node propagation path and record the results. Finally, the generated directional constrained diffusion channel not only shows the interaction between the various parts, but also provides optimization suggestions to help identify and solve potential problems in the design.

[0123] 303. Adjusting the path connectivity priority of the topological connection probability field according to the distribution gradient of the spatial attenuation weight to generate a constrained diffusion network extending from a high attenuation area to a low attenuation area;

[0124] In step 303, path connectivity prioritization adjusts the order of the topological connection probability field based on the distribution gradient of the spatial attenuation weights. High-attenuation areas are prioritized, followed by low-attenuation areas. This prioritization ensures that critical areas of the design (such as emergency evacuation routes) are prioritized while maintaining the rationality and functionality of the overall layout. A constrained diffusion network, a network structure that extends from high-attenuation areas to low-attenuation areas, prioritizes critical components of the design while maintaining the consistency and coherence of the overall design.

[0125] In an embodiment of the present application, first, the spatial attenuation weight is mapped to a spatial gradient field, and the gradient amplitude and direction of each node are calculated. Next, a directed connectivity graph is constructed based on the gradient direction, and the adjacent gradient edges pointing to the current node with a gradient amplitude lower than the current node are retained. Then, the normalized inverse of the gradient amplitude is superimposed as the path selection weight to generate an attenuation-aware connection priority score. Finally, the nodes are traversed from high to low according to the connection priority score to generate a constrained diffusion network extending from the high attenuation area to the low attenuation area, ensuring that the key areas are given priority.

[0126] 304. Perform hierarchical clustering on the constrained diffusion network based on the constrained diffusion channel, establish cross-hierarchical transmission links through overlapping nodes, and form a constraint relationship diffusion network.

[0127] In step 304, hierarchical clustering is the process of categorizing and organizing the constraint diffusion network. Cross-level transmission links are established through overlapping nodes, integrating information from different levels. This method helps optimize the overall layout and ensure that all design elements meet preset standards. The constraint relationship diffusion network, the final network structure formed through hierarchical clustering, demonstrates the interactions between various components and provides comprehensive design optimization recommendations, ensuring that the design solution is both aesthetically pleasing and practical.

[0128] In the present embodiment, the generated constraint diffusion network is first clustered hierarchically to identify overlapping nodes. Next, cross-level transmission links are established to integrate information from different levels. The model is then run to verify the validity and connectivity of the paths. Finally, the resulting constraint relationship diffusion network not only demonstrates the interactions between the various components but also provides comprehensive design optimization recommendations, ensuring that the design solution is both aesthetically pleasing and practical.

[0129] In summary, steps 301 to 304, through precise data analysis and intelligent optimization, seamlessly transitioned architectural design from concept to actual implementation. This significantly enhanced the feasibility and aesthetics of the design proposal, while also significantly improving construction quality and efficiency. This ensured that every detail accurately reflected the designer's original intention, resulting in a building that was both visually striking and highly functional, particularly for emergency situations. This approach not only addressed complex design challenges but also optimized the overall layout and enhanced the user experience.

[0130] The step 303 of adjusting the path connectivity priority of the topological connection probability field according to the distribution gradient of the spatial attenuation weight to generate a constrained diffusion network extending from the high attenuation area to the low attenuation area includes:

[0131] 401. Map the spatial attenuation weight into a spatial gradient field, where the gradient amplitude of each node is the norm of the weight difference of adjacent nodes, and the gradient direction is the neighborhood direction where the weight decreases fastest;

[0132] In step 401, the spatial gradient field is the result of mapping the spatial attenuation weights into a virtual space. The gradient magnitude of each node represents the norm of the weight difference between adjacent nodes (i.e., the difference in the rate of change of curvature), while the gradient direction points to the direction of the neighborhood where the weight decreases most rapidly. This structure helps identify key areas that require priority in the design. The gradient magnitude refers to the absolute value of the weight difference between each node and its adjacent nodes, which measures the importance of the node and its influence on surrounding nodes. The gradient direction describes the direction of the neighborhood where the weight decreases most rapidly, guiding how to expand the path from high-attenuation areas to low-attenuation areas.

[0133] In the embodiment of the present application, first, the spatial attenuation weight is mapped to a spatial gradient field, and the gradient magnitude and direction of each node are calculated. Then, a mathematical algorithm is used to determine that the gradient direction of each node is the neighborhood direction where the weight decreases the fastest. Then, this information is used to construct a preliminary spatial gradient field to ensure that the gradient magnitude of each node accurately reflects its importance. Finally, this data is integrated to form a detailed spatial gradient field, which provides a basis for subsequent steps.

[0134] 402. Construct a directed connected graph based on the gradient direction, retain for each node only the adjacent gradient edges pointing to a node with a gradient magnitude lower than that of the current node, and filter out the reverse gradient edges;

[0135] In step 402, a directed connectivity graph is constructed based on gradient directions. For each node, only adjacent gradient edges with a lower gradient magnitude than the current node are retained, and reverse gradient edges are filtered out. This approach ensures path validity and connectivity. In a directed connectivity graph, a gradient edge is a directional connection from one node to another. Only edges with a lower gradient magnitude than the current node are retained.

[0136] In the embodiment of the present application, a directed connectivity graph is first constructed based on the gradient direction, retaining the adjacent gradient edges pointing to the node with a lower gradient magnitude than the current node. Next, all reverse gradient edges are filtered out to ensure the validity and connectivity of the path. Then, the model is run to verify the validity of the path and the results are recorded. Finally, the resulting directed connectivity graph provides clear path guidance for subsequent steps, ensuring that key areas are prioritized.

[0137] 403. Generate a decay-aware connection priority score based on the initial connection probability of each node in the topological connection probability field and the normalized inverse of the gradient amplitude as a path selection weight.

[0138] In step 403, the path selection weight is calculated by superimposing the normalized inverse of the gradient amplitude of each node's initial connection probability in the topological connection probability field. This approach ensures that path selection not only considers the connectivity probability of functional intervals but also the impact of curvature changes. The decay-aware connection priority score, generated from the path selection weight, reflects the priority of each node in path selection. Nodes with higher scores are selected first, ensuring that the design meets physical and functional requirements.

[0139] In the embodiment of the present application, first, based on the initial connection probability of each node in the topological connection probability field, the normalized inverse of the gradient amplitude is superimposed as the path selection weight. Next, these weights are used to generate a decay-aware connection priority score. The model is then run to evaluate the effectiveness of the path selection and the results are recorded. Finally, the generated connection priority score provides a clear basis for path selection in subsequent steps, ensuring that the design solution is both aesthetically pleasing and practical.

[0140] 404. Traverse the nodes from high to low according to the connection priority score, starting from the seed node with the maximum spatial attenuation weight, dynamically expand the connection edges along the gradient descent direction, and generate a constrained diffusion network extending from the high attenuation area to the low attenuation area.

[0141] In step 404, the seed node, the starting node with the largest spatial attenuation weight, is used to initiate the path extension process. The selection of seed nodes is crucial to the entire path extension process. Dynamically extending connecting edges is a process that starts from the seed node and gradually expands the path along the gradient descent direction. This method ensures that the path extends from high-attenuation areas to low-attenuation areas, optimizing the overall layout.

[0142] In the embodiment of the present application, first, starting from the seed node with the maximum spatial attenuation weight, the connection edge is dynamically expanded along the gradient descent direction. Then, the nodes are traversed from high to low according to the connection priority score, and the path is gradually expanded. Then, the model is run to verify the effectiveness of the path expansion and the results are recorded. Finally, the generated constrained diffusion network extending from the high attenuation area to the low attenuation area not only shows the interaction between the various parts, but also provides comprehensive design optimization suggestions to ensure that the design scheme is both beautiful and practical.

[0143] In summary, steps 401 to 404, through precise data analysis and intelligent optimization, seamlessly transition architectural design from concept to implementation. This significantly enhances the feasibility and aesthetics of the design, while also significantly improving construction quality and efficiency. The resulting building is both visually striking and highly functional, particularly for emergency situations. This approach not only addresses complex design challenges but also optimizes the overall layout and enhances the user experience. The generation of a constrained diffusion network ensures that the design meets both aesthetic and functional requirements, particularly in crowd management and emergency evacuation.

[0144] The topological relationship verification of the functional area connection node parameters and the functional area boundary constraint conditions in step 104 to generate a building scene adaptive correction parameter set includes:

[0145] 501. Perform parameter decomposition of the real-time curvature gradient parameters in manifold space, establish an orthogonal parameter coordinate system along the main curvature direction of the surface, and convert the continuous gradient change into a discretized morphological deviation tensor;

[0146] In step 501, the real-time curvature gradient parameter refers to a data set that describes the rate of change of surface curvature. Parametric decomposition of manifold space simplifies complex geometric shapes into a form that can be analyzed and processed. An orthogonal parametric coordinate system is a reference frame consisting of two mutually perpendicular coordinate axes that is used to define position and orientation on a surface. The morphological deviation tensor is a quantitative representation of the difference between the actual shape of a surface and its ideal shape. These data work together to achieve a precise description of the shape of a complex surface.

[0147] In the embodiment of the present application, a specific algorithm is first used to perform a manifold space parameter decomposition on the acquired real-time curvature gradient parameters to simplify the surface model. Next, an orthogonal parameter coordinate system is constructed according to the main curvature direction to facilitate subsequent calculations. Then, through this coordinate system, the continuous gradient changes are converted into a discretized morphological deviation tensor to more accurately capture the surface details. Finally, based on the data obtained by the above process, a preliminary quantitative description of the surface shape is completed.

[0148] 502. Establish a dynamic constraint network of the functional area connection node parameters through a spatial topological relationship parameter library, and generate a surface morphology correction field according to the morphology deviation tensor and the functional area boundary constraint conditions;

[0149] In step 502, the spatial topology parameter library is a data set that stores various possible spatial connection patterns and their corresponding constraints. A dynamic constraint network is a network model that automatically adjusts its structure based on input parameters. The morphological deviation tensor and functional region boundary constraints serve as inputs to guide the direction and degree of surface morphological correction. The surface morphological correction field is a set of specific values ​​that indicate how the surface should be adjusted.

[0150] In the embodiment of the present application, a dynamic constraint network is first established based on the relevant information extracted from the spatial topology parameter library. Next, the morphological deviation tensor and the functional area boundary constraints are used as input to calculate the required correction amount for each node through an optimization algorithm. These correction amounts are then integrated to form a surface morphology correction amount field. Ultimately, this amount field is used to guide subsequent correction work.

[0151] 503. Inject the function area connection node parameters into the dynamic constraint network, perform topological energy diffusion calculation based on the constraint propagation path of the function area connection nodes, and generate a topological correction gradient distribution cloud map;

[0152] In step 503, the constraint propagation path refers to the way information or constraints are transmitted within the network. Topological energy diffusion calculation is a method for simulating the distribution of energy within a system. The topological correction gradient distribution cloud map is a map used to visually display the intensity of correction requirements at each point.

[0153] In the present embodiment, the parameters of the functional area connection nodes are first injected into the established dynamic constraint network. Next, information is transmitted according to the preset constraint propagation path. Then, a topological energy diffusion calculation method is used to evaluate the energy state of each node and generate a topological correction gradient distribution cloud map. Finally, the optimal correction strategy is determined by analyzing this cloud map.

[0154] 504. Establish a joint optimization space of the surface morphology correction field and the topology correction gradient distribution cloud map, construct a potential energy surface of the joint action of morphology and topology through dual-parameter coupling, perform parameter space correction along the direction of potential energy gradient descent, and generate a building scene adaptive correction parameter set.

[0155] In step 504, the joint optimization space refers to a multi-dimensional parameter space that simultaneously considers both morphological and topological characteristics. Dual-parameter coupling involves adjusting two different parameters simultaneously to achieve the optimal solution. The potential energy surface of the combined morphological and topological effects is a mathematical model that reflects the mutual influence between morphological and topological characteristics. The potential energy gradient descent direction seeks the path that minimizes the potential energy.

[0156] In this embodiment, a joint optimization space is first established between the surface morphology correction field and the topology correction gradient distribution cloud. Next, a dual-parameter coupling technique is used to explore the optimal solution within this space. Parameters are then adjusted along the potential energy gradient descent until the most appropriate correction solution is found. Finally, a parameter set for adaptive correction of the architectural scene is generated, containing all necessary adjustment information.

[0157] In summary, steps 501 to 504 significantly improve the ability to process complex curved surfaces in the architectural design process, making the transition between functional areas more natural and smooth, while also enhancing the safety and aesthetics of the overall structure.

[0158] The step 504 of establishing a joint optimization space of the surface morphology correction field and the topology correction gradient distribution cloud map, and constructing a potential energy surface of the joint action of morphology and topology through dual-parameter coupling, includes:

[0159] 601. Perform spatial discretization processing on the surface morphology correction field to generate a gridded morphology correction parameter matrix covering the building scene surface, and convert the topology correction gradient distribution cloud map into a node gradient intensity tensor of the same spatial resolution;

[0160] In step 601, the gridded morphology correction parameter matrix is ​​a set of parameters generated by spatially discretizing the surface morphology correction field. It contains the displacement correction components (Δx, Δy, Δz) for each grid node in three-dimensional space. The node gradient strength tensor is a matrix of scalar values ​​of the topological constraint strength corresponding to each node after converting the topology correction gradient distribution cloud map to the same grid resolution. The surface morphology correction field stores the deformation adjustment for each point on the building surface, and the topology correction gradient distribution cloud map reflects the spatial distribution strength of the constraint conditions during the structural topology optimization process.

[0161] In the embodiment of the present application, the finite element meshing technology is first used to discretize the building surface into a quadrilateral or triangular unit grid, and each grid vertex is used as a parameter node. Then, the spatial gradient distribution data of the surface morphological correction field is extracted through the deformation gradient calculation module to generate a parameter matrix containing the correction values ​​in the three directions of x / y / z. The topological correction gradient distribution cloud map is then processed by a bilinear interpolation algorithm, and its original resolution is adjusted to be consistent with the morphological correction grid, forming a gradient intensity value in the range of 0-1 corresponding to each node. Finally, the spatial position of the morphological correction parameters and the topological gradient parameters is ensured by alignment through the three-dimensional space coordinate system.

[0162] 602. Establish a spatial mapping relationship between the gridded morphology correction parameter matrix and the node gradient intensity tensor, and form a joint parameter space through parameter coordinate registration;

[0163] In step 602, the spatial mapping relationship refers to the mathematical correspondence rules between the morphological correction parameters and the topological gradient parameters in spatial positions; the joint parameter space is a multidimensional parameter set space formed by coordinate alignment of the morphological correction parameter matrix and the gradient intensity tensor, which contains the deformation parameters and topological constraint strength combination data of each node.

[0164] In the embodiment of the present application, a coordinate system conversion model is first established between the morphological correction grid and the topological gradient grid, and an affine transformation algorithm is used to eliminate rotational and translational deviations between the two grids. Next, a bidirectional association mapping table is established between each morphological correction node and the topological gradient node at the corresponding position through node index matching technology. A five-dimensional parameter space data structure is then constructed containing node coordinates, morphological correction vectors, and topological gradient values. Finally, a spatial topological relationship verification algorithm is used to detect and repair mapping anomalies in the grid boundary area.

[0165] 603. Define a two-variable coupling relationship between the surface morphology correction field and the topology correction gradient distribution cloud in the joint parameter space, and construct a two-parameter coupling relationship reflecting the interaction between morphological deformation energy and topological constraint energy;

[0166] In step 603, the two-variable coupling relationship refers to the interaction between the morphological correction factor and the topological constraint. This two-parameter coupling relationship is a mathematical model used to describe how these two variables interact to achieve an optimal solution. The morphological deformation energy reflects the energy required to change shape, while the topological constraint energy represents the energy barrier required to maintain a specific connection pattern. Together, these terms reveal the intrinsic connection between morphology and topology.

[0167] In the embodiments of the present application, a bivariate coupling relationship between the surface morphology correction field and the topology correction gradient distribution cloud is first defined in a joint parameter space. Next, an optimization algorithm is used to determine the optimal bivariate coupling relationship that reflects the interaction between morphological deformation energy and topological constraint energy. The validity of this relationship is then verified through simulation and calculation. Finally, through this process, a mathematical model that closely links morphology and topology is established.

[0168] 604. Perform a full differential expansion on the dual-parameter coupling relationship, combine the surface curvature continuity and topological constraint propagation characteristics, and construct a potential energy surface of the joint action of morphology and topology.

[0169] In step 604, the total differential expansion is a mathematical method used to analyze the changing trends of complex functions. Curvature continuity refers to the smoothness of the curvature changes at each point on a surface. The topological constraint propagation property describes how topological conditions are propagated throughout the system. The potential energy surface of the combined morphology and topology is a visualization tool that shows the interaction between morphology and topology. These concepts work together to guide design decisions.

[0170] In the embodiment of the present application, the dual-parameter coupling relationship is first fully differentially expanded to gain a deeper understanding of its changing pattern. Next, a potential energy surface of the combined action of morphology and topology is constructed by combining the surface curvature continuity and the propagation characteristics of topological constraints. This potential energy surface is then visualized using computer-aided design software to intuitively evaluate the effects of various design solutions. Finally, through this series of steps, a design solution that comprehensively considers morphology and topology factors is obtained.

[0171] In summary, steps 601 to 604, by establishing a model combining morphological correction and topological constraints, achieve collaborative design that optimizes building surface deformation and improves structural performance. In a commercial complex project, this method effectively addressed the difficult balance between aesthetic form and structural efficiency in irregularly shaped curved buildings, reducing stress concentration at key nodes by over 30%, increasing construction error tolerance by 50%, and optimizing the overall solution's economic performance by over 18%.

[0172] The step 105 of reconstructing the processing trajectory control points of the building components based on the building scene adaptive correction parameter set to generate the building design scene construction instructions including the linkage between spatial form and functional topology includes:

[0173] 701. Decompose the building scene adaptive correction parameter set into a morphological correction parameter matrix and a topological correction gradient tensor, wherein the morphological correction parameter matrix includes three-dimensional displacement components of surface control points;

[0174] In step 701, the architectural scene adaptive correction parameter set is a set of data derived from previous analysis to guide architectural form and topology optimization. The form correction parameter matrix is ​​a data structure that describes the three-dimensional displacement of surface control points, while the topology correction gradient tensor is a data structure that quantitatively describes the changing trends of topological constraints. These concepts work together to achieve a precise description of architectural form and internal topological relationships.

[0175] In the embodiment of the present application, the correction parameter set is first decomposed into two independent data streams, morphology and topology, through a data decoupling module. The morphology correction parameter matrix is ​​stored in a three-dimensional tensor structure, with each element containing a triplet of control point number, original coordinates, and correction vector data. The topology correction gradient tensor is converted into a relative intensity value matrix in the range of 0-1 through normalization. Finally, a spatial index comparison table is established for the two sets of data to ensure that the node numbers strictly correspond to the physical locations.

[0176] 702. Performing spatiotemporal discretization processing on the morphology correction parameter matrix based on a parameterized coordinate system, and generating an initial distribution of control points of the machining trajectory according to bicubic spline interpolation;

[0177] In step 702, a parametric coordinate system is a reference frame that defines how to mathematically represent the position of an object. Spatiotemporal discretization is the process of converting continuous spatial and temporal information into a discrete set of data points. Bicubic spline interpolation is a technique for smooth curve fitting. These concepts work together to generate the initial distribution of control points for the machining trajectory.

[0178] In the embodiments of the present application, a spatiotemporal discretization process is first performed on the morphology correction parameter matrix based on an established parameterized coordinate system. Next, bicubic spline interpolation is applied to generate an initial distribution of control points for the machining trajectory based on the processed data. This initial distribution is then evaluated and adjusted to ensure that it meets the actual machining requirements. Finally, this process lays the foundation for subsequent control point optimization.

[0179] 703. Map the topology correction gradient tensor to the space where the control points of the machining trajectory are initially distributed, and apply topology correction weights to the control points along the gradient association channel to form a composite control point field.

[0180] In step 703, topology correction gradient tensor mapping refers to converting topological constraints into a data format that can be used to guide control point adjustment. The composite control point field is a spatial environment that integrates morphological correction and topological correction information. The gradient correlation channel is the path used to transmit topological correction information. These terms work together to achieve precise adjustment of control points.

[0181] In the present embodiment, a three-dimensional spatial attenuation model of the topological gradient is first constructed, where the gradient intensity propagates exponentially along the component force transmission path. Next, a gradient-associated channel is searched for within the initial control point distribution, and a triangulation is used to establish the spatial topological relationship between the control points. A weighting factor ω is then calculated based on the position of the control point in the gradient channel, and the initial control point coordinates are weighted and adjusted. Ultimately, a composite data field is generated, comprising the original coordinates, the correction vector, and the gradient weight.

[0182] 704. Establish a trajectory optimization model in the composite control point field, eliminate trajectory mutations between control points through local curvature adaptive adjustment, and transfer the topology correction gradient to adjacent control point clusters to build a hierarchical control network;

[0183] In step 704, the trajectory optimization model is a mathematical model used to find the optimal machining path. Local curvature adaptive adjustment dynamically adjusts the path based on local surface features. The hierarchical control network assigns different levels of management strategies based on the importance of different control points. These data work together to construct an efficient machining path.

[0184] In the present embodiment, the moving least squares method is first used to construct a local surface and calculate the curvature change rate between control points. Auxiliary control points are then inserted into the sudden change section, and the trajectory is refitted using a quintic polynomial curve. A control point influence attenuation model is then constructed based on the gradient tensor, with the gradient influence weight of the core node set to 0.8, the secondary node to 0.5, and the base node to 0.2. Finally, a hierarchical association relationship is established using a breadth-first search algorithm to form a tree-like control network structure.

[0185] 705. Spatially couple the control point cluster of the key feature area with the topology correction gradient to generate construction instructions for an architectural design scenario that includes a linkage between spatial form and functional topology.

[0186] In step 705, key feature areas refer to parts of a building structure that have special functions or aesthetic value. A control point cluster is a set of interconnected control points used to accurately describe geometric changes within a specific area. A topological correction gradient is a data structure that quantitatively describes the changing trends of topological constraints. Spatial coupling refers to the process of combining different types of parameters or data sets through spatial relationships.

[0187] In an embodiment of the present application, the key feature areas in the architectural design are first identified, and their corresponding control point clusters are determined. Next, based on the topological correction gradient generated in the previous step, the spatial adjustment amount required for the control point clusters in each key feature area is calculated. Then, spatial coupling technology is used to combine the topological correction gradient with the specific location information of the control point cluster to ensure that each key feature area meets both the overall morphological requirements and the local functional requirements. Finally, based on the above processing results, construction instructions for the architectural design scene that include the linkage between spatial morphology and functional topology are generated to provide clear guidance for actual construction.

[0188] In summary, steps 701 to 705 achieve the coordinated optimization of building component processing trajectories and structural performance through the coupled control of morphology and topology. This significantly improves the ability to handle complex morphological and topological relationships during the architectural design process, not only optimizing the building's exterior but also enhancing the rationality and safety of the interior space layout. It also improves the accuracy and operability of construction instructions, effectively facilitating the smooth implementation of the project.

[0189] Figure 2 The present invention provides a schematic diagram of a system for automatically generating architectural design scenes based on artificial intelligence. Figure 2 As shown, the system includes:

[0190] An acquisition module 21 acquires a geometric feature parameter library of special-shaped components including a surface curvature gradient change threshold and a spatial topological relationship parameter library including functional area boundary constraint conditions of the architectural design scene;

[0191] An analysis module 22 performs a joint analysis of the surface curvature gradient change threshold and the functional area boundary constraint conditions through an artificial intelligence optimization engine to generate a dynamic morphological constraint map of the building scene;

[0192] Extraction module 23, which acquires three-dimensional space lattice data of the construction site in real time through laser point cloud scanning, and extracts real-time curvature gradient parameters and functional area connection node parameters from the three-dimensional space lattice data;

[0193] Verification module 24 performs spatial morphological matching on the real-time curvature gradient parameter and the surface curvature gradient change threshold in the dynamic morphological constraint map of the building scene, and simultaneously verifies the topological relationship between the functional area connection node parameter and the functional area boundary constraint condition, thereby generating a building scene adaptive correction parameter set;

[0194] The generation module 25 reconstructs the processing trajectory control points of the building components based on the building scene adaptive correction parameter set, and generates the building design scene construction instructions including the linkage between spatial form and functional topology.

[0195] Figure 2 The artificial intelligence-based automatic generation system for architectural design scenes can be executed Figure 1 The implementation principles and technical effects of the AI-based automated generation method for architectural design scenarios described in the illustrated embodiment are not further elaborated. The specific manner in which each module and unit performs operations in the AI-based automated generation system for architectural design scenarios described in the aforementioned embodiment has been described in detail in the relevant embodiments of the method and will not be further elaborated here.

[0196] In one possible design, Figure 2 The embodiment shown is an artificial intelligence-based automatic generation system for architectural design scenes that can be implemented as a computing device, such as Figure 3 As shown, the computing device may include a storage component 31 and a processing component 32;

[0197] The storage component 31 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component 32 .

[0198] The processing component 32 is used for the above Figure 1 The embodiment provides an artificial intelligence-based automatic generation method for architectural design scenes.

[0199] The processing component 32 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented as one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above method.

[0200] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0201] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. An artificial intelligence-based automatic generation method for architectural design scenes, characterized in that: include: Obtaining a geometric feature parameter library of special-shaped components including a surface curvature gradient change threshold and a spatial topological relationship parameter library including functional area boundary constraint conditions for architectural design scenarios; The artificial intelligence optimization engine is used to jointly analyze the surface curvature gradient change threshold and the functional area boundary constraint conditions to generate a dynamic morphological constraint map of the building scene; Acquire three-dimensional space dot matrix data of the construction site in real time through laser point cloud scanning, and extract real-time curvature gradient parameters and functional area connection node parameters from the three-dimensional space dot matrix data; Performing spatial morphological matching between the real-time curvature gradient parameter and the surface curvature gradient change threshold in the dynamic morphological constraint map of the building scene, and performing topological relationship verification between the functional area connection node parameter and the functional area boundary constraint condition, to generate a building scene adaptive correction parameter set; Reconstructing the processing trajectory control points of the building components based on the building scene adaptive correction parameter set to generate construction instructions for the building design scene that include the linkage between spatial form and functional topology; The artificial intelligence optimization engine is used to jointly analyze the surface curvature gradient change threshold and the functional area boundary constraint condition to generate a dynamic morphological constraint map of the building scene, including: Decomposing the surface curvature gradient change threshold into a curvature gradient feature vector, and decoupling the functional area boundary constraint condition into a multi-dimensional topological constraint feature vector; Mapping the curvature gradient eigenvector and the multidimensional topological constraint eigenvector to the same coupling analysis space, and generating a hybrid constraint tensor that links geometric form and spatial topology through tensor fusion; Constructing a dynamic constraint diffusion model based on the hybrid constraint tensor, driving the node propagation path of the multidimensional topological constraint feature vector with the spatial attenuation characteristics of the curvature gradient feature vector to form a constraint relationship diffusion network; The constraint relationship diffusion network is reversely corrected and its boundaries are extracted. A hierarchical dynamic morphological constraint boundary is generated according to the spatial continuity of the curvature gradient sensitive domain and the functional topological conduction chain, and then reorganized into a dynamic morphological constraint map of the architectural scene.

2. The method according to claim 1, characterized in that The step of driving the node propagation path of the multi-dimensional topological constraint feature vector using the spatial attenuation characteristic of the curvature gradient feature vector to form a constraint relationship diffusion network includes: Converting the curvature gradient eigenvector in the hybrid constraint tensor into a spatial attenuation weight matrix, and converting the multidimensional topological constraint eigenvector into a topological connection probability field; Dynamically weighting the topological connection probability field based on the spatial attenuation weight matrix to generate a directional constrained diffusion channel, wherein the activation intensity of the constrained diffusion channel is dynamically adjusted by the relationship between the spatial attenuation coefficient and the topological connection probability; Adjusting the path connectivity priority of the topological connection probability field according to the distribution gradient of the spatial attenuation weight to generate a constrained diffusion network extending from a high attenuation area to a low attenuation area; The constrained diffusion network is hierarchically clustered based on the constrained diffusion channel, and cross-hierarchical conduction links are established through overlapping nodes to form a constrained relationship diffusion network.

3. The method according to claim 2, characterized in that The step of adjusting the path connectivity priority of the topological connection probability field according to the distribution gradient of the spatial attenuation weight to generate a constrained diffusion network extending from a high attenuation area to a low attenuation area includes: The spatial attenuation weight is mapped to a spatial gradient field, where the gradient amplitude of each node is the norm of the weight difference of the adjacent nodes, and the gradient direction is the neighborhood direction where the weight decreases fastest; Constructing a directed connected graph based on the gradient direction, retaining for each node only the adjacent gradient edges pointing to nodes with a gradient magnitude lower than the current node, and filtering the reverse gradient edges; According to the initial connection probability of each node in the topological connection probability field, the normalized inverse of the gradient amplitude is superimposed as a path selection weight to generate a decay-aware connection priority score; The nodes are traversed from high to low according to the connection priority scores, and starting from the seed node with the maximum spatial attenuation weight, the connection edges are dynamically expanded along the gradient descent direction to generate a constrained diffusion network extending from the high attenuation area to the low attenuation area.

4. The method according to claim 1, wherein The topological relationship verification of the functional area connection node parameters and the functional area boundary constraint conditions to generate a building scene adaptive correction parameter set includes: The real-time curvature gradient parameters are parameterized in manifold space, an orthogonal parameter coordinate system is established along the main curvature direction of the surface, and the continuous gradient changes are converted into discretized morphological deviation tensors; A dynamic constraint network of the functional area connection node parameters is established through a spatial topological relationship parameter library, and a surface morphology correction field is generated according to the morphological deviation tensor and the functional area boundary constraint conditions; Injecting the functional area connection node parameters into the dynamic constraint network, performing topological energy diffusion calculation based on the constraint propagation path of the functional area connection node, and generating a topological correction gradient distribution cloud map; A joint optimization space of the surface morphology correction field and the topology correction gradient distribution cloud map is established, and a potential energy surface of the joint action of morphology and topology is constructed through dual-parameter coupling. The parameter space is corrected along the descending direction of the potential energy gradient to generate an adaptive correction parameter set for the building scene.

5. The method according to claim 4, characterized in that The method of establishing a joint optimization space of the surface morphology correction field and the topology correction gradient distribution cloud map, and constructing a potential energy surface of the joint action of morphology and topology through dual-parameter coupling, includes: The surface morphology correction field is spatially discretized to generate a gridded morphology correction parameter matrix covering the building scene surface, and the topology correction gradient distribution cloud map is converted into a node gradient intensity tensor of the same spatial resolution; Establishing a spatial mapping relationship between the gridded morphology correction parameter matrix and the node gradient intensity tensor, and forming a joint parameter space through parameter coordinate registration; Defining a two-variable coupling relationship between the surface morphology correction field and the topology correction gradient distribution cloud in the joint parameter space, and constructing a two-parameter coupling relationship reflecting the interaction between morphological deformation energy and topological constraint energy; The two-parameter coupling relationship is fully differentially expanded, and the surface curvature continuity and topological constraint propagation characteristics are combined to construct the potential energy surface of the joint action of morphology and topology.

6. The method according to claim 1, characterized in that The process of reconstructing the processing trajectory control points of the building components based on the building scene adaptive correction parameter set to generate the building design scene construction instructions including the linkage between spatial form and functional topology includes: Decomposing the building scene adaptive correction parameter set into a morphological correction parameter matrix and a topological correction gradient tensor, wherein the morphological correction parameter matrix includes three-dimensional displacement components of surface control points; Performing spatiotemporal discretization processing on the morphological correction parameter matrix based on a parameterized coordinate system, and generating an initial distribution of control points of the machining trajectory according to bicubic spline interpolation; Mapping the topology correction gradient tensor to the space where the control points of the machining trajectory are initially distributed, applying topology correction weights to the control points along the gradient association channel to form a composite control point field; A trajectory optimization model is established in the composite control point field, trajectory mutations between control points are eliminated through local curvature adaptive adjustment, and topology correction gradients are transferred to adjacent control point clusters to construct a hierarchical control network; The control point clusters of the key feature areas are spatially coupled with the topology correction gradient to generate construction instructions for architectural design scenarios that include the linkage between spatial form and functional topology.

7. An artificial intelligence-based automatic generation system for architectural design scenes, used in the artificial intelligence-based automatic generation method for architectural design scenes according to any one of claims 1 to 6, characterized in that: include: An acquisition module is used to obtain a geometric feature parameter library of special-shaped components including a surface curvature gradient change threshold and a spatial topological relationship parameter library including functional area boundary constraint conditions of the architectural design scene; An analysis module, which uses an artificial intelligence optimization engine to jointly analyze the surface curvature gradient change threshold and the functional area boundary constraint conditions to generate a dynamic morphological constraint map of the building scene; An extraction module acquires three-dimensional space dot matrix data of the construction site in real time through laser point cloud scanning, and extracts real-time curvature gradient parameters and functional area connection node parameters from the three-dimensional space dot matrix data; A verification module performs spatial morphological matching on the real-time curvature gradient parameter and the surface curvature gradient change threshold in the dynamic morphological constraint map of the building scene, and simultaneously verifies the topological relationship between the functional area connection node parameter and the functional area boundary constraint condition, thereby generating a building scene adaptive correction parameter set; A generation module reconstructs the processing trajectory control points of the building components based on the building scene adaptive correction parameter set, and generates construction instructions for the building design scene that include the linkage between spatial form and functional topology.

8. A computing device, characterized in that It includes a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement an artificial intelligence-based automatic generation method for architectural design scenes as described in any one of claims 1 to 6.

9. A computer storage medium, characterized in that A computer program is stored, and when the computer program is executed by a computer, the method for automatically generating an architectural design scene based on artificial intelligence as described in any one of claims 1 to 6 is implemented.

Images