Three-dimensional voxel-based spatial design optimization method, system, terminal and storage medium

Abstract

The application relates to the technical field of building information analysis, and discloses a three-dimensional voxel-based space design optimization method, system, terminal and storage medium, the method comprising the following steps: determining three-dimensional coordinates according to a building planning scheme, converting the three-dimensional coordinates into a voxel grid, and obtaining voxel codes of each three-dimensional coordinate; acquiring rule information, constructing voxels corresponding to the rule information on the voxel codes, and weighting the rule information according to compliance rules; determining voxel density of the voxel codes according to the weight values, predicting potential problems at the voxel codes, and thus adjusting the building planning scheme in real time. Through the three-dimensional voxel-based space analysis process, multi-source data is fused and weighted through voxels, automatic optimization of each kind of data is realized, and high-precision, dynamic, multi-target space analysis and scheme optimization are realized.

Description

Technical Field

[0001] This invention relates to the field of building information analysis technology, and in particular to a spatial design optimization method, system, terminal, and computer-readable storage medium based on three-dimensional voxelization. Background Technology

[0002] Current spatial analysis technologies in various industries mainly rely on traditional two-dimensional drawing analysis, static three-dimensional modeling, and single-dimensional data processing methods.

[0003] For example, in the field of architectural engineering, designers often use CAD (Computer Aided Design) to draw two-dimensional drawings and combine them with simple three-dimensional models for spatial layout planning; in urban planning, GIS (Geographic Information System) is often used for geographic information overlay analysis, but it is difficult to achieve deep integration and dynamic adjustment of multi-source data; in the field of geology and mining, borehole data interpolation modeling is often used, which has weak real-time analysis and conflict detection capabilities for underground space.

[0004] Therefore, existing technologies still need to be improved and developed. Summary of the Invention

[0005] The main objective of this invention is to provide a spatial design optimization method, system, terminal, and computer-readable storage medium based on three-dimensional voxelization, aiming to solve the problems of poor dynamic adaptability, low efficiency, and single analysis dimension in existing spatial analysis technologies, which cannot meet the current industry's demand for high-precision, dynamic, and full-scene spatial analysis.

[0006] To achieve the above objectives, the present invention provides a spatial design optimization method based on three-dimensional voxelization, the spatial design optimization method based on three-dimensional voxelization comprising the following steps:

[0007] Obtain the architectural planning scheme for the target area, determine multiple geodetic coordinates in the geodetic coordinate system based on the architectural planning scheme, and map all the geodetic coordinates to a three-dimensional voxel coordinate system to obtain multiple three-dimensional coordinates of the target area;

[0008] All the three-dimensional coordinates are converted into corresponding voxel grids to obtain the voxel code of each three-dimensional coordinate in the voxel grid;

[0009] Obtain various rule information of the target region, construct voxels corresponding to all data information in each rule information on each voxel encoding, and assign weights to each data information according to the defined compliance rules;

[0010] Based on the weight of each data information, determine multiple voxel densities for all voxel codes, and based on all voxel densities, predict the detection result of the potential problem at the corresponding voxel code.

[0011] The building planning scheme is adjusted in real time based on all the test results, and the weights of all data information in all the rule information are updated in real time to generate a construction plan for the target area.

[0012] In this invention, an architectural planning scheme for a target area is obtained. Based on this scheme, multiple geodetic coordinates are determined in a geodetic coordinate system. All geodetic coordinates are mapped to a three-dimensional voxel coordinate system to obtain multiple three-dimensional coordinates of the target area. All three-dimensional coordinates are converted into corresponding voxel grids to obtain voxel codes for each coordinate. Multiple rule information for the target area is obtained. Voxels corresponding to all data information within each rule information are constructed on each voxel code, and weights are assigned to each data information according to defined compliance rules. Multiple voxel densities of all voxel codes are determined based on the weights of each data information, and the detection results of potential problems at the corresponding voxel codes are predicted based on all voxel densities. The architectural planning scheme is adjusted in real-time based on all detection results to update the weights of all data information within all rule information in real-time, generating a construction plan for the target area. This invention, through a three-dimensional voxelized spatial analysis process, fuses and weights multi-source data via voxel fusion, achieving automatic optimization of each data type and realizing high-precision, dynamic, and multi-objective spatial analysis and scheme optimization. Attached Figure Description

[0013] Figure 1 This is a flowchart of a preferred embodiment of the spatial design optimization method based on three-dimensional voxelization of the present invention;

[0014] Figure 2 This is a structural diagram of a preferred embodiment of the spatial design optimization system based on three-dimensional voxelization of the present invention;

[0015] Figure 3 This is a structural diagram of a preferred embodiment of the terminal of the present invention. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this invention clearer and more explicit, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0017] The spatial design optimization method based on three-dimensional voxelization described in the preferred embodiment of the present invention, such as... Figure 1As shown, the spatial design optimization method based on three-dimensional voxelization includes the following steps:

[0018] Step S10: Obtain the architectural planning scheme of the target area, determine multiple geodetic coordinates in the geodetic coordinate system according to the architectural planning scheme, and map all the geodetic coordinates to the three-dimensional voxel coordinate system to obtain multiple three-dimensional coordinates of the target area.

[0019] The geodetic coordinates include: geodetic abscissa, geodetic ordinate, and geodetic vertical coordinate. Based on the local coordinate system of the target area, it needs to be converted to the geodetic coordinate system through a unified transformation standard. In one embodiment of the present invention, the parameters of the geodetic coordinate system can adopt the CGCS2000 ellipsoid (major semi-axis a=6378137m, flattening f=1 / 298.257222101). The local coordinate system parameters of the target area need to obtain the ellipsoid parameters, projection method, and seven transformation parameters of the local independent coordinate system. These data can be confirmed by the local surveying and mapping department or project surveying data.

[0020] Specifically, the architectural planning scheme of the target area is obtained, and all boundary points of the architectural planning scheme are converted into geodetic abscissa, corresponding geodetic ordinate, and corresponding geodetic vertical coordinate in the geodetic coordinate system using a seven-parameter transformation method; Gaussian projection is performed on all the geodetic abscissa and all the geodetic ordinate respectively to obtain the corresponding planar abscissa and corresponding planar vertical coordinate; a three-dimensional voxel coordinate system is constructed according to the target area, and each of the planar abscissa, each of the planar vertical coordinate, and each of the geodetic vertical coordinates is projected into the three-dimensional voxel coordinate system to obtain the three-dimensional coordinates of each point in the target area.

[0021] Furthermore, based on the above embodiments, the lowest point in the southwest corner of the project area (minimum plane coordinates and minimum elevation) is used as the origin of the voxel coordinate system. It is ensured that all voxel coordinates are non-negative. The absolute coordinates of the origin must be recorded simultaneously in both geodetic 2000 coordinates (x0_2000, y0_2000, z0_2000) and local coordinates for subsequent coordinate mapping with external systems. The x, y, and z axes of the three-dimensional voxel coordinate system are set parallel to the x, y, and z axes of the geodetic coordinate system, respectively. Through the above process, the mapping deviation between the three-dimensional voxel coordinate system and the geodetic 2000 coordinate system must be ≤0.05m. Then, more than 10 evenly distributed control points (possessing both geodetic 2000 and voxel coordinates) are periodically selected for verification. If the deviation exceeds the limit, it is corrected using a "coordinate deviation correction algorithm".

[0022] Step S20: Convert all the three-dimensional coordinates into corresponding voxel grids to obtain the voxel code of each three-dimensional coordinate in the voxel grid.

[0023] The voxel grid includes equidistant grid lines for the horizontal coordinate, the vertical coordinate, and the vertical axis. After the transformation of the three-dimensional voxel coordinates is realized, the target area is automatically divided into a continuous voxel grid through image rendering, and an editable three-dimensional voxel model is generated and exported in multiple formats to improve the applicability of the model. Then, each voxel is assigned a unique voxel code to ensure accurate mapping of spatial position.

[0024] Specifically, based on the type of the current construction stage of the target area, the corresponding voxel size is determined; based on the voxel size, the three-dimensional voxel coordinate system is divided into equidistant grid lines for the horizontal coordinate, the vertical coordinate, and the elongated grid lines; based on each three-dimensional coordinate, a unique identifier for each three-dimensional coordinate is determined on the equidistant grid lines for the horizontal coordinate, the vertical coordinate, and the elongated grid lines; based on each unique identifier, each point is encoded to obtain the corresponding voxel code.

[0025] The partitioning process is automated by the program based on input data and preset rules, eliminating the need for manual specification of key parameters such as partitioning range and voxel dimensions. First, it automatically reads project boundary data (such as closed polylines in CAD, polygon vector data in GIS, and site boundaries from BIM (Building Information Modeling) models), and identifies boundary vertex coordinates using a "vector boundary extraction algorithm" (supporting multiple mainstream formats). For example, it reads the polyline from the "project boundary" layer in a CAD file and extracts the geodetic 2000 coordinates (x, y, z) of its 100 vertices from (X1, Y1, Z1) to (X100, Y100, Z100). Then, it automatically calculates the maximum and minimum coordinate values ​​of the target area to determine the bounding box of the entire 3D space (including the x, y, and z axes), and generates three-dimensional bounding box parameters for the target area (including not only the maximum and minimum coordinate values ​​but also the parameters of all points within the entire target area), serving as the partitioning benchmark.

[0026] Furthermore, based on the project type and accuracy requirements, a voxel size calculation model is invoked. This model determines different accuracy coefficients for different projects (identified from the metadata of the input file), and then calculates the voxel size based on these coefficients. For example, for "urban planning," the accuracy coefficient is set to 1000, and the voxel side length is the difference between the maximum and minimum coordinate values ​​divided by 1000. Different accuracy coefficients are adaptively selected to control the voxel size according to different scene types, preventing excessively large voxels from causing insufficient accuracy.

[0027] Furthermore, based on the voxels corresponding to different points, corresponding attribute data (such as weight at the corresponding point, building material, and building compliance) needs to be generated and stored in the database using coordinate codes as index keys. In the quality inspection after mesh generation (such as checking for duplicate voxels), the uniqueness of the coordinate codes can quickly identify duplicate codes, thereby greatly improving the inspection efficiency.

[0028] Furthermore, in the process of discretization and voxel generation in three-dimensional space, "division" is the process of decomposing a continuous three-dimensional space into regular voxel units. In another embodiment of this invention, a process based on the spatial discretization capabilities of the OpenGL (Open Graphics Library) graphics engine is disclosed. Specifically, every three adjacent X, Y, and Z-axis grid lines enclose a voxel unit. A "triple loop algorithm" traverses all grid line intersections to generate voxels; then, the three coordinate axes are looped to assign a unique coordinate code to each voxel. The OpenGL engine stores the coordinates of the voxel's eight vertices using a Vertex Buffer Object (VBO) and defines the voxel's six faces (each face consisting of four vertices) using an Index Buffer Object (IBO), thus realizing the graphical construction of voxels.

[0029] Furthermore, when the length of one axis of the project bounding box is not an integer multiple of the voxel size, the last voxel may exceed the boundary. The system automatically executes the clipping algorithm to ensure that the voxel does not exceed the project range and adds an identifier after the coordinate code (such as the coordinate code "X101C-Y050-Z00", with the identifier "C" added), marking it as a clipped voxel, which will be specially handled in subsequent analysis (such as adjusting the weight according to the actual volume ratio when calculating the weight).

[0030] During the partitioning process, OpenGL renders the voxel mesh in real time (wireframe or solid mode), and users can visually observe the partitioning progress. If any abnormality is found, the process can be manually terminated (but it is fully automatic by default).

[0031] Format Export Adaptation: When generating the preset format, OpenGL writes information such as voxel coordinate codes and spatial ranges into extended fields through vertex attribute pointers, ensuring that the software can read the voxel mesh structure and attribute associations. After partitioning, the system automatically performs three checks: continuity between adjacent voxels, integrity of each voxel, and whether the coordinate codes of all voxels are duplicated. If a problem is found, automatic correction is triggered: vertex coordinates of discontinuous voxels are recalculated, missing areas are partitioned, and duplicate codes are reassigned.

[0032] When the project area contains voxels of multiple scales (such as a 0.5m voxel in the core area and a 1m voxel in the periphery), the coordinate encoding achieves seamless splicing through "scale identifier extension". For example, the encoding format of a 0.5m voxel is "X001-0-Y002-0-Z003-0" (the 0 at the end indicates a 0.5m scale), corresponding to a spatial range of 0.0~0.5m (X-axis); the 1m voxel encoding "X001-Y002-Z003" can be associated with two 0.5m voxels ("X001-0 / Y002-0 / Z003-0" and "X001-1 / Y002-0 / Z003-0"), realizing the logical association of grids of different scales through encoding rules.

[0033] Furthermore, when exporting the voxel mesh to a universal format after generation, coordinate encoding automatically generates absolute coordinates recognizable by external systems through the "encoding-absolute coordinate transformation formula." For example: Absolute X coordinate (Geometry 2000) = X0_2000 + voxel × sequence number × voxel size. Through the above process, the exported mesh model can be accurately positioned in different external systems during the conversion process, enabling spatial overlay analysis with other project data (such as terrain and pipelines). The coordinate system construction standard ensures accurate mapping between the 3D voxel coordinate system and Geometry 2000 and local coordinate systems by clearly defining the datum transformation, origin definition, and precision control. Coordinate encoding, as the "digital ID card" of voxels, plays a core role in voxel mesh generation, including positioning, indexing, multi-scale adaptation, and external interaction, and is a key technical means to achieve accurate mesh division and efficient management.

[0034] Step S30: Obtain multiple rule information of the target region, construct voxels corresponding to all data information in each rule information on each voxel encoding, and assign weights to each data information according to the defined compliance rules.

[0035] The rules include geographic information, functional information, and real-time monitoring data. Each data point is initially authorized based on user-input weight settings. When external conditions change (such as adjustments to height restrictions or changes in functional requirements due to population growth), the system automatically retrieves relevant data (such as the latest documents and population statistics) and updates the corresponding voxel weights. For example, if the height restriction in a certain area is adjusted from 50m to 80m, the compliance weight of voxels in the original 50-80m height restriction range automatically updates from 0 to 1, without manual intervention.

[0036] Specifically, the system acquires various environmental information from the geographic information of the target area, and constructs a voxel corresponding to each type of environmental information at each voxel encoding; acquires various demand information from the functional information of the target area, and constructs a voxel corresponding to each type of demand information at each voxel encoding; acquires various monitoring results from the real-time monitoring data of the target area, and constructs a voxel corresponding to each type of monitoring result at each voxel encoding; acquires compliance rules input by the user, calculates the matching degree between each type of environmental information, each type of demand information, and each type of monitoring result and the compliance rules, and assigns weights to each type of environmental information, each type of demand information, and each type of monitoring result based on all the matching degrees.

[0037] This involves acquiring various types of information about the target area, such as geographic information, functional requirements information, and real-time monitoring data. These information include multiple types of data, each with its own weight and represented by different voxels. Each type of data can be represented in different formats to improve the system's applicability.

[0038] Furthermore, the compliance rules input by the user are obtained, and the compliance information input by the user is converted into multiple voxel attribute thresholds; according to each voxel attribute threshold, compliance checks are performed on each type of environmental information, each type of demand information, and each type of monitoring result at each voxel encoding, to obtain matching values ​​corresponding to all environmental information, all demand information, and all monitoring results at each voxel encoding; according to all the matching values, each type of environmental information, each type of demand information, and each type of monitoring result is assigned weights respectively.

[0039] In one embodiment of the present invention, geographic information is accessed in the form of images, including slope, sunshine duration, etc. The weights corresponding to these two are the weights of terrain difficulty and livability, respectively. The conversion process can be achieved through terrain slope classification and sunshine intensity quantification.

[0040] The process involves radiometric correction of the images, calculation of the average annual sunshine duration for each pixel, calculation of the noon solar angle value (maximum value in summer, minimum value in winter, average value is taken during calculation) in conjunction with the project latitude (i.e., the latitude of the target area), and calculation of the comprehensive quantification value using a weighted summation method (e.g., sunshine duration accounts for 70%, solar altitude angle accounts for 30%). The relationship between different elements and weights is automatically matched (e.g., the weights corresponding to different slopes, the weights corresponding to different sunshine levels, i.e., the quantification process of both). Moreover, these weights are not fixed, but are adjusted in real time according to the actual situation.

[0041] For example, in cold regions, the sunshine weight needs to be 20% higher than that of ordinary building areas, and in mountainous regions, the slope weight needs to be 30% higher than that of ordinary building areas. For each new project, the system can predict the most weighted value through machine learning based on historical project data (e.g., a correlation model of "weight-construction cost" composed of multiple completed projects), thereby automatically adjusting the weights of each element of the current project. It also supports experts manually inputting weight correction coefficients (range 0.5~1.5), with the corrected weight = baseline value × correction coefficient. For example, if an expert judges that a certain plot of land has a higher actual development difficulty, inputting a correction coefficient of 1.2 will adjust the "steep slope" weight from 0.8 to 0.96. Both terrain slope grading and sunshine intensity quantification have clear standards to ensure the standardization of data processing; the example weights are only initial baselines, dynamically adjusted through a four-tiered mechanism of "project adaptation + regional adaptation + algorithm optimization + manual intervention" to ultimately meet the precise assignment needs of different scenarios before mapping to voxels.

[0042] In another embodiment of the present invention, the intervention of functional requirement information can be accessed in the form of multiple parameters, specifically including different functional areas (commercial area, warehousing area, etc.). The floor area ratio of the functional area is converted into functional priority, thereby obtaining the weight of different functional areas. Therefore, the weight calculation process can be converted into the priority calculation process.

[0043] The plot ratio does not directly map to priority. Instead, it is linked to the functional zoning through the logic of "basic plot ratio of functional zoning → deviation analysis of actual plot ratio of the project → priority correction". The core is to reflect the impact of "whether the plot ratio matches the functional requirements" on the value of space.

[0044] First, based on different functional zones and in accordance with the target area's conversion rules (including international and local standards), different benchmark floor area ratios are determined for each functional zone. Then, using a two-level benchmark floor area ratio database, benchmark values ​​can be automatically matched. If the project has special planning requirements (such as free trade zones), the benchmark floor area ratio approved by the planning department can be manually entered.

[0045] Then, the actual plot ratio deviation rate is calculated to quantify the difference. The deviation rate is calculated using the following formula based on different parameters:

[0046] Floor area ratio deviation rate = (Actual floor area ratio - Benchmark floor area ratio) / Benchmark floor area ratio × 100%;

[0047] If the deviation rate is greater than 0, the actual floor area ratio is higher than the benchmark, which may be due to over-development, so the priority needs to be reduced; if the deviation rate is less than 0, the actual floor area ratio is lower than the benchmark, which indicates underdevelopment, so the priority needs to be increased; if the deviation rate is equal to 0, the current priority of this functional area does not need to be modified.

[0048] Finally, based on the range of deviation rates, the baseline priority of the functional partitions is corrected to form the final functional priority. The system automatically calculates the deviation rate and matches the correction coefficient. If the corrected priority exceeds 1.0, it is set to 1.0 (upper limit); if it is below 0.1, it is set to 0.1 (lower limit), ensuring that the priority is within a reasonable range.

[0049] Furthermore, real-time monitoring data represents the detection results from the equipment, and each detection result can be assigned different weights according to its importance.

[0050] The core of functional zoning priority lies in the contribution of each functional zone to the current project in the target area. This process requires dual evaluation through industry standards and project-specific requirements to determine the priority of each functional zone. The plot ratio is used to link functional zoning into priorities via a process of "benchmark value → deviation rate → correction coefficient," quantifying the rationality of development. Functional zoning is scored using the AHP (Analytic Hierarchy Process) algorithm across four dimensions: economic, social, objective, and feasibility, combined with industry standards and project positioning to determine the baseline priority. The commercial priority of 0.9 represents a balance between "core value + risk control + planning requirements," reflecting the core position of commerce while reserving space for dynamic adjustments and functional equilibrium.

[0051] Step S40: Determine multiple voxel densities for all voxel codes based on the weights of each type of data information, and predict the detection results of potential problems at the corresponding voxel codes based on all voxel densities.

[0052] After assigning different weights to various data points in the current project, conflict detection can be performed based on the building rules of the target area. This includes spatial conflict detection, compliance conflict detection, and resource conflict detection, with two detection frequencies: real-time and scheduled detection. Real-time detection is used during the design phase, triggering a detection immediately upon modification of each voxel parameter. Scheduled detection (e.g., once per hour) can be set during the construction and operation phases to balance detection accuracy and system energy consumption.

[0053] Specifically, all environmental information, all demand information, and all monitoring results with weights not lower than preset values ​​are retained as voxels to obtain the voxel densities corresponding to the geographic information, functional information, and real-time monitoring data at each voxel encoding. For each voxel encoding, the geographic information, functional information, and real-time monitoring data are clustered to extract geographic information tags, functional information tags, and real-time monitoring tags. Based on the extracted geographic information tags, functional information tags, and real-time monitoring tags, the degree of density exceeding the standard and the time dimension of density exceeding the standard for the voxel densities of the geographic information, functional information, and real-time monitoring data are determined respectively. Based on all the degree of density exceeding the standard and the corresponding time dimension, all detection results at the voxel encoding are predicted.

[0054] Specifically, spatial conflict detection involves comparing the spatial relationships between voxels in real time, identifying overlapping relationships between building component voxels (beams, columns, pipelines, etc.) and intersections between voxels of lower ore bodies and roadways, thus preventing collisions from occurring.

[0055] For compliance conflict detection, a compliance rule engine needs to be established to convert the ecological red line rules into voxel attribute threshold standards and compare them with the constructed voxels to make a judgment. In another embodiment of the present invention, the vector boundary data of the ecological red line is obtained from the GIS spatial database and projected into a three-dimensional voxel coordinate system. For a voxel, if its centroid coordinates fall within the spatial range enclosed by the vector boundary of the ecological red line, the voxel is determined to be in the ecological red line area. For example, the ecological red line boundary is determined by a series of latitude and longitude coordinates, which are converted into a coordinate range (minimum x-coordinate, maximum x-coordinate, minimum y-coordinate, maximum y-coordinate, minimum vertical coordinate, maximum vertical coordinate) (Xmin, Xmax, Ymin, Ymax, Zmin, Zmax) in the voxel coordinate system through coordinate system transformation. When the centroid coordinates (x, y, z) of a voxel satisfy Xmin≤Xv≤Xmax, Ymin≤Yv≤Ymax, and Zmin≤Zv≤Zmax, the voxel is marked as being within the ecological red line, and the corresponding compliance weight threshold is set to 0.

[0056] Furthermore, in some complex situations, a single voxel may encompass multiple land uses, some of which lie within ecological red lines. In this case, the proportion of the area within the ecological red line to the total area of ​​the voxel is calculated. If this proportion exceeds a certain threshold (e.g., 50%, which can be adjusted according to local ecological protection requirements), the voxel is determined to be primarily located within the ecological red line area, and the compliance weight threshold of 0 is applied. For example, by performing spatial overlay analysis on the ecological red line polygon within the voxel and the boundary polygon of the voxel, the area of ​​the overlapping portion is calculated, and then divided by the total area of ​​the voxel to obtain the proportion, which is used as the basis for judgment.

[0057] Furthermore, the compliance rules also include height restriction rules (which include dynamic height restriction adjustments, such as at airports). In a voxel coordinate system, the top elevation of each voxel is compared with the height restriction value. If a voxel exceeds the height restriction threshold, it is immediately marked as a compliance conflict. For areas with dynamic height restriction requirements, such as height restriction areas around airports that change with distance, or height restriction adjustments during specific periods (such as when hosting major events), the system dynamically updates the height restriction threshold of voxels based on the height restriction rules acquired in real time. By establishing a functional relationship between height restriction and spatial location (such as distance from the airport runway center) and time, the system calculates the height restriction threshold of each voxel in the current situation in real time and performs compliance checks. For example, different height restriction values ​​are set within different radii around the airport with the runway center as the center. Voxels calculate the corresponding height restriction threshold based on their spatial location and perform height compliance judgment.

[0058] Each voxel is assigned a corresponding land use code. If the land use code of a voxel does not match the code required by the planning regulations, it is considered a land use conflict. For example, if a planned area is residential land (code R2), and the land use code of a voxel is commercial land (code B1), then the voxel exceeds the land use threshold and is marked as a compliance conflict. In practice, mixed land use exists; a land use proportion threshold is set to determine the dominant land use. For example, if a voxel contains both residential and commercial land uses, and the residential land area accounts for more than 70% (which can be adjusted according to local planning), then the dominant land use of the voxel is determined to be residential; if the commercial land area accounts for more than 70%, then the dominant land use is commercial. If the dominant land use does not conform to the planning requirements, it is also marked as a compliance conflict. By calculating the area and analyzing the proportion of polygons with different land uses within a voxel, the dominant land use is determined, and compliance is judged.

[0059] Furthermore, the core service objectives, resource consumption types, and user behavior characteristics of high-weight functional voxels are fundamentally different. Therefore, the potential problems caused by exceeding the preset threshold are completely different. For example, when the density of commercial voxels exceeds the standard, the focus is on the problem of "matching people flow with consumption resources," which may cause queuing and congestion. When the density of transportation hub voxels exceeds the standard, the focus is on the problem of "matching traffic flow with carrying capacity," which may cause queuing and delays. When the density of medical facility voxels exceeds the standard, the focus is on the problem of "matching service capacity with patient needs," which may cause queuing and resource shortages. When the density of educational facility voxels exceeds the standard, the focus is on the problem of "matching teaching resources with student size," which may cause overcrowded classes and insufficient activity space.

[0060] Therefore, in another embodiment of the present invention, the distribution density of high-weight functional voxels (such as commercial and transportation hubs) is analyzed based on the K-means clustering algorithm. When the voxel clustering density of a certain area exceeds a preset threshold (such as commercial voxel density > 5 / 100 square meters), it is determined to be a resource conflict, and potential problems (such as pedestrian congestion and insufficient parking) are predicted.

[0061] Different functional voxels have different core service requirements. When the density exceeds the standard, the first criterion for judgment is to determine the scope of the problem based on the voxel type. See Table 1 below for details:

[0062] Table 1: Problem Category Mapping Table

[0063]

[0064] After K-means clustering is completed, the type label of each voxel is automatically read (such as "commercial-shopping mall" and "transportation-subway station"), and the corresponding potential problem categories are matched to eliminate irrelevant problems (such as when the density of commercial voxels exceeds the standard, there is no need to judge "bed shortage").

[0065] Furthermore, based on the threshold exceeded by voxel density, the severity and urgency of potential problems are determined. This process requires assessment using the exceedance rate.

[0066] Exceedance rate = (Actual cluster density - Preset threshold) / Preset threshold × 100%;

[0067] Taking the commercial voxels (threshold 5 voxels / 100 square meters) in the above embodiment as an example, if the density exceeds the standard by less than 20%, it indicates a slight exceedance. Since the flow of people is concentrated in specific areas (such as the food and beverage floor), the overall capacity still has redundancy. Therefore, the potential problem is local congestion during peak hours (such as queuing in the food and beverage area). If the density exceeds the standard by 20%-50%, it indicates a moderate exceedance. Since the flow density of people in the main passage is >1.5 people / square meter (the upper limit of the standard), and the waiting time for service facilities is >10 minutes, the potential problem is local congestion throughout the day and queuing for some service facilities (such as cash registers and elevators). If the density exceeds the standard by more than 50%, it indicates a serious exceedance. Since the flow density of people in the core passage is >2.5 people / square meter, and the fire evacuation time exceeds the standard requirements (such as >5 minutes), the potential problem is large-scale congestion throughout the day and safety risks (such as the risk of stampede).

[0068] Furthermore, since excessive density is only a symptom, the core problem is the mismatch between the service demand of functional elements and the supply capacity of surrounding supporting resources. Therefore, it is necessary to identify specific resource gaps by linking the supporting resource database and clarify the specific manifestations of potential problems.

[0069] In another embodiment of the present invention, a transportation hub voxel (subway station, density 3.5 voxels / 100 square meters, exceeding the standard rate 16.7%) is used as an example for illustration:

[0070] The supporting data for this clustered area was extracted from the supporting resource database. This included two entrances / exits for pedestrian areas with a standard requirement of more than three per 1000m², main passageways with a standard requirement of more than three meters in width (2m wide), and four elevators, each with a service capacity of less than 500 people / hour. Then, comparing demand and supply, the target area required a pedestrian flow of 2000 people / hour (peak hours), which translates to a supply of 4 x 500 elevators, a perfect match. However, the supply of entrances / exits and passageways was insufficient. Insufficient entrances / exits led to slow pedestrian evacuation, and insufficient passageway width caused congestion. Therefore, the potential problem was identified as: "Congestion during peak hours due to insufficient entrances / exits." To address this issue, a database linking supporting resources and functional voxels needs to be established, storing the supporting resource types (e.g., transportation hub → entrances / exits, passageways, elevators) and supply capacity parameters for each functional voxel. A demand-supply comparison algorithm needs to be developed to automatically calculate the resource gap rate (gap rate = (demand - supply) / supply × 100%). Resource items with a gap rate > 0 are identified as potential problem areas.

[0071] Furthermore, when determining compliance rules, the potential problems corresponding to density exceeding the standard in the same functional voxel clustering area at different times are different (e.g., commercial voxels during the day and night, weekdays and holidays), requiring further precise judgment based on time-series data. In another embodiment disclosed in this invention, commercial voxels (density 5.8 voxels / 100 square meters, exceeding the standard rate of 16%) are used for illustration:

[0072] First, the density change curve of the target area is extracted from the time-series density database to determine the time period when the density exceeds the standard. Then, combined with the time-series data of user behavior, the specific reasons for the exceedance are determined (i.e., what common user behaviors led to the peak). Finally, the predicted potential problems are output: the potential problem on weekdays is crowded queues in the catering area, and the corresponding solution is to temporarily add more seats; the potential problem on holidays is queues at the cash registers in the retail area and congestion at the parking lot entrance, and the corresponding solutions are to temporarily add cash registers and add guides at the parking lot entrance.

[0073] This invention collects temporal density data (sampling every 15 minutes) and user behavior data (such as consumption type and dwell time) from functional voxel clustering regions, and trains a time-problem association model (such as a decision tree model). Inputting temporal density and time labels, it automatically outputs potential problems for the corresponding time period. The core logic is to first define the broad category (functional type) → then classify it into levels (exceeding the standard rate) → then identify pain points (supporting resources) → finally determine the timeliness (time dimension). Through this four-dimensional framework, it can delve into the essential problem from the apparent density, avoiding the simplistic attribution of all density exceeding the standard to "congestion," thus achieving accuracy and practicality in resource conflict detection.

[0074] Step S50: Adjust the building planning scheme in real time based on all the detection results, and update the weight of all data information in all the rule information in real time to generate a construction plan for the target area.

[0075] In this invention, users can customize the multi-objective optimization weights and use an improved non-dominated sorting genetic algorithm for optimization. In another embodiment of the invention, the population size can be set to 50, the number of iterations to 100, the crossover probability to 80, and the mutation probability to 0.05. The system can automatically search for the optimal solution for the multi-objective and finally output an automatically generated solution, which includes core indicators (space utilization, sunshine duration, cost, compliance rate) and optimization suggestions.

[0076] Specifically, resource allocation is performed based on all the detection results and the time dimension of all density exceeding the standard, and all the environmental information, all the demand information, and all the monitoring results are updated in real time; according to the compliance rules, the weights of all the current environmental information, all the demand information, and all the monitoring results are updated in real time, and the voxels corresponding to the geographical information, the functional information, and the real-time monitoring data at each voxel encoding are updated according to all the weights; based on all the updated voxels, a construction plan for the target area is generated.

[0077] The following is an example of an embodiment of the present invention based on the above-described embodiments: a residential project of 100,000 square meters.

[0078] The core requirements are space utilization rate ≥90%, sunlight satisfaction rate ≥85%, and construction cost ≤12,000 yuan / square meter, generating 5 feasible solutions. A 3D voxel model of the project area has been completed, with voxels measuring 1m×1m×1m, totaling 100,000 voxels. Of these, 70% are residential function voxels, 20% are supporting facility voxels, and 10% are public space voxels. Based on the above algorithm parameters, the target weights are optimized, setting space utilization rate to 0.4, sunlight satisfaction rate to 0.3, and construction cost to 0.3.

[0079] First, the core design parameters of the residential project are converted into chromosome codes that the algorithm can recognize, using real number encoding (this effectively avoids the precision loss of binary encoding). Each code corresponds to an individual scheme, as shown in Table 2 below:

[0080] Table 2: Encoding Structure Table

[0081]

[0082] Then, Latin hypercube sampling was used to generate 50 initial scheme individuals to ensure that the population was evenly distributed in the solution space and to avoid the initial schemes being concentrated. For example, individual 1 was coded as [70, 65, 25, 20] (residential density 70 units / 1000 square meters, building height 65m, building spacing 25m, supporting facilities ratio 20%). For each initial scheme individual, its fitness value was calculated under three objectives: space utilization rate, sunshine satisfaction rate, and construction cost (quantifying the merits of the scheme). Finally, the overall fitness was calculated based on the three fitness values ​​and their corresponding weights.

[0083] Furthermore, the non-dominated sorting algorithm, the core of the genetic algorithm, is used to divide the 50 initial scheme individuals into different levels according to their dominance relationship. For example, if all target fitness values ​​of scheme A are greater than or equal to those of scheme B, and at least one target fitness value is greater than that of scheme B, then A dominates B. The final sorting result is that the first level is the optimal level, with no other scheme dominating individuals (a total of 8 individuals), the second level consists of individuals dominated only by the first level (a total of 12 individuals), and the third level and below consist of individuals dominated by the first two levels (a total of 30 individuals). This process prioritizes the retention of higher-level individuals to ensure that the algorithm evolves towards the optimal solution.

[0084] Furthermore, 80% (32 individuals) of the individuals from levels 1 and 2 are randomly selected for crossover. New individuals are generated with a crossover probability of 0.8. Then, parameters are mutated for all individuals (including original individuals and crossover individuals) with a mutation probability of 0.05 to avoid the algorithm getting trapped in local optima. The parameter mutation magnitude is set to be less than 10% of the original parameters to ensure the feasibility of the scheme. After each iteration, the top 50 individuals with the highest overall fitness (original + crossover + mutated) from the parent generation are retained as the next generation population, while individuals with lower fitness are eliminated. Finally, from the individuals in level 1 of the 100th iteration, 5 schemes are selected based on goal balance to ensure that each scheme has its own advantages on different goals. This invention utilizes an improved genetic algorithm to efficiently complete multi-objective optimization, providing a scientific and feasible design scheme while significantly improving the efficiency of scheme generation. It reduces invalid iterations through non-dominated sorting and "elite retention" processes, and only performs calculations on high-fitness individuals, reducing computational power consumption. All five schemes obtained meet the core requirements of "space utilization ≥ 90%, sunshine satisfaction ≥ 85%, and cost ≤ 12,000 yuan / square meter". Among them, scheme 4 (space 92.0%, sunshine 87.8%, cost 11,600 yuan / square meter) is the most balanced among the three objectives and is selected as the final recommended scheme.

[0085] This invention achieves high-precision, dynamic, and multi-objective spatial analysis and scheme optimization by fusing and weighting multi-source data through a three-dimensional voxelization spatial analysis process.

[0086] Furthermore, such as Figure 2 As shown, based on the above-described spatial design optimization method based on three-dimensional voxelization, the present invention also provides a spatial design optimization system based on three-dimensional voxelization, wherein the spatial design optimization system based on three-dimensional voxelization includes:

[0087] The three-dimensional voxelization module 51 is used to obtain the architectural planning scheme of the target area, determine multiple geodetic coordinates in the geodetic coordinate system according to the architectural planning scheme, and map all the geodetic coordinates to the three-dimensional voxel coordinate system to obtain multiple three-dimensional coordinates of the target area.

[0088] The annotation module 52 is used to convert all the three-dimensional coordinates into corresponding voxel grids to obtain the voxel code of each three-dimensional coordinate in the voxel grid;

[0089] The weight assignment module 53 is used to obtain various rule information of the target region, construct voxels corresponding to all data information in each rule information on each voxel encoding, and assign weights to each data information according to the defined compliance rules.

[0090] The multi-source data fusion module 54 is used to determine multiple voxel densities of all voxel codes according to the weight of each type of data information, and predict the detection result of potential problems at the corresponding voxel codes according to all voxel densities.

[0091] The scheme update module 55 is used to adjust the building planning scheme in real time based on all the detection results, so as to update the weight of all data information in all the rule information in real time and generate the construction scheme of the target area.

[0092] Furthermore, such as Figure 3 As shown, based on the above-mentioned spatial design optimization method and system based on three-dimensional voxelization, the present invention also provides a terminal, which includes a processor 10, a memory 20 and a display 30. Figure 3 Only some of the terminal components are shown; however, it should be understood that it is not required to implement all of the components shown, and more or fewer components may be implemented instead.

[0093] In some embodiments, the memory 20 may be an internal storage unit of the terminal, such as a hard disk or memory. In other embodiments, the memory 20 may be an external storage device of the terminal, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc. Further, the memory 20 may include both internal and external storage devices. The memory 20 is used to store application software and various types of data installed on the terminal, such as the program code installed on the terminal. The memory 20 can also be used to temporarily store data that has been output or will be output. In one embodiment, the memory 20 stores a three-dimensional voxel-based spatial design optimization program 40, which can be executed by the processor 10 to implement the three-dimensional voxel-based spatial design optimization method of this application.

[0094] In some embodiments, the processor 10 may be a central processing unit (CPU), a microprocessor, or other data processing chip, used to run program code stored in the memory 20 or process data, such as executing the spatial design optimization method based on three-dimensional voxelization.

[0095] In some embodiments, the display 30 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen. The display 30 is used to display information on the terminal and to display a visual user interface. The components of the terminal communicate with each other via a system bus.

[0096] In one embodiment, when the processor 10 executes the three-dimensional voxelization-based spatial design optimization program 40 in the memory 20, it implements the steps of the three-dimensional voxelization-based spatial design optimization method as described above.

[0097] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a spatial design optimization program based on three-dimensional voxelization, and the spatial design optimization program based on three-dimensional voxelization, when executed by a processor, implements the steps of the spatial design optimization method based on three-dimensional voxelization as described above.

[0098] In summary, this invention provides a spatial design optimization method and related equipment based on three-dimensional voxelization. The method includes: acquiring an architectural planning scheme for a target area; determining multiple geodetic coordinates in a geodetic coordinate system based on the architectural planning scheme; mapping all the geodetic coordinates to a three-dimensional voxel coordinate system to obtain multiple three-dimensional coordinates of the target area; converting all the three-dimensional coordinates into corresponding voxel grids to obtain voxel codes for each three-dimensional coordinate in the voxel grids; acquiring multiple rule information of the target area; constructing voxels corresponding to all data information in each rule information on each voxel code; assigning weights to each data information according to defined compliance rules; determining multiple voxel densities of all voxel codes based on the weights of each data information; predicting the detection results of potential problems at the corresponding voxel codes based on all voxel densities; and adjusting the architectural planning scheme in real time based on all the detection results to update the weights of all data information in all the rule information in real time, thereby generating a construction plan for the target area. This invention achieves high-precision, dynamic, and multi-objective spatial analysis and scheme optimization by fusing and weighting multi-source data through a three-dimensional voxelization spatial analysis process.

[0099] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal that includes that element.

[0100] Of course, those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware (such as a processor, controller, etc.). The program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The computer-readable storage medium can be a memory, magnetic disk, optical disk, etc.

[0101] It should be understood that the application of the present invention is not limited to the examples above. Those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A spatial design optimization method based on three-dimensional voxelization, characterized in that, The spatial design optimization method based on three-dimensional voxelization includes: Obtain the architectural planning scheme for the target area, determine multiple geodetic coordinates in the geodetic coordinate system based on the architectural planning scheme, and map all the geodetic coordinates to a three-dimensional voxel coordinate system to obtain multiple three-dimensional coordinates of the target area; All the three-dimensional coordinates are converted into corresponding voxel grids to obtain the voxel code of each three-dimensional coordinate in the voxel grid; Obtain various rule information of the target region, construct voxels corresponding to all data information in each rule information on each voxel encoding, and assign weights to each data information according to the defined compliance rules; The rule information includes: geographic information, functional information, and real-time monitoring data; The process of acquiring multiple rule information for the target region, constructing voxels corresponding to all data information in each rule information on each voxel encoding, and assigning weights to each data information according to defined compliance rules specifically includes: Obtain various environmental information from the geographic information of the target area, and construct a voxel corresponding to each type of environmental information at each voxel encoding location; Obtain multiple requirement information from the functional information of the target region, and construct a voxel corresponding to each requirement information at each voxel encoding location; Acquire multiple monitoring results from the real-time monitoring data of the target area, and construct a voxel corresponding to each monitoring result at each voxel encoding; Obtain the compliance rules input by the user, calculate the degree of matching between each type of environmental information, each type of demand information, and each type of monitoring result and the compliance rules, and assign weights to each type of environmental information, each type of demand information, and each type of monitoring result based on all the degree of matching; Based on the weight of each data information, determine multiple voxel densities for all voxel codes, and based on all voxel densities, predict the detection result of the potential problem at the corresponding voxel code. The building planning scheme is adjusted in real time based on all the test results, and the weights of all data information in all the rule information are updated in real time to generate a construction plan for the target area.

2. The spatial design optimization method based on three-dimensional voxelization according to claim 1, characterized in that, The process of obtaining the architectural planning scheme for the target area, determining multiple geodetic coordinates in a geodetic coordinate system based on the architectural planning scheme, and mapping all the geodetic coordinates to a three-dimensional voxel coordinate system to obtain multiple three-dimensional coordinates of the target area, specifically includes: Obtain the architectural planning scheme for the target area, and convert all boundary points of the architectural planning scheme into the geodetic abscissa, the corresponding geodetic ordinate, and the corresponding geodetic ordinate in the geodetic coordinate system using the seven-parameter transformation method; Gaussian projection is performed on all the said geodetic abscissas and all the said geodetic ordinates respectively to obtain the corresponding planar abscissas and the corresponding planar ordinates; A three-dimensional voxel coordinate system is constructed based on the target region. Each plane horizontal coordinate, each plane vertical coordinate, and each geodetic vertical coordinate are projected onto the three-dimensional voxel coordinate system to obtain the three-dimensional coordinates of each point in the target region.

3. The spatial design optimization method based on three-dimensional voxelization according to claim 2, characterized in that, The voxel grid includes: equidistant grid lines for the horizontal axis, equidistant grid lines for the vertical axis, and equidistant grid lines for the vertical axis. The step of converting all the three-dimensional coordinates into corresponding voxel grids to obtain the voxel code of each three-dimensional coordinate in the voxel grid specifically includes: Determine the corresponding voxel size based on the type of the current construction phase of the target area; Based on the voxel size, the three-dimensional voxel coordinate system is divided into equidistant grid lines for the horizontal coordinate, equidistant grid lines for the vertical coordinate, and equidistant grid lines for the vertical axis. Based on each of the three-dimensional coordinates, determine a unique identifier for each of the three-dimensional coordinates on the equidistant grid lines of the horizontal coordinate, the equidistant grid lines of the vertical coordinate, and the equidistant grid lines of the vertical axis. Each point is encoded based on its unique identifier to obtain the corresponding voxel code.

4. The spatial design optimization method based on three-dimensional voxelization according to claim 1, characterized in that, The process of obtaining compliance rules input by the user, calculating the degree of matching between each type of environmental information, each type of demand information, and each type of monitoring result and the compliance rules, and assigning weights to each type of environmental information, each type of demand information, and each type of monitoring result based on all the matching degrees, specifically includes: Obtain compliance rules input by the user and convert the compliance information input by the user into various voxel attribute thresholds; Based on each of the voxel attribute thresholds, compliance checks are performed on each of the environmental information, each of the demand information, and each of the monitoring results at each voxel encoding location to obtain matching values ​​corresponding to all environmental information, all demand information, and all monitoring results at each voxel encoding location. Based on all the said matching values, each of the said environmental information, each of the said demand information, and each of the said monitoring results are assigned weights respectively.

5. The spatial design optimization method based on three-dimensional voxelization according to claim 1, characterized in that, The step of determining multiple voxel densities for all voxel codes based on the weights of each type of data information, and predicting the detection result of potential problems at the corresponding voxel codes based on all voxel densities, specifically includes: All environmental information, all demand information and all monitoring results with weights not lower than preset values ​​are retained as voxels to obtain the voxel density corresponding to the geographic information, the functional information and the real-time monitoring data at each voxel encoding. For each voxel code, the geographic information, the functional information, and the real-time monitoring data are clustered to extract geographic information tags, functional information tags, and real-time monitoring tags; Based on the extracted geographic information tags, the functional information tags, and the real-time monitoring tags, the degree of density exceeding the standard and the time dimension of density exceeding the standard of the voxel density of the geographic information, the functional information, and the real-time monitoring data are determined respectively. Based on all the aforementioned density exceedance levels and the corresponding time dimension, predict all detection results at the voxel encoding.

6. The spatial design optimization method based on three-dimensional voxelization according to claim 5, characterized in that, The step of adjusting the building planning scheme in real time based on all the detection results, updating the weights of all data information in all the rule information in real time, and generating a construction plan for the target area specifically includes: Resource allocation will be carried out based on all the test results and all time dimensions of density exceeding the standard, and all the environmental information, all the demand information and all the monitoring results will be updated in real time. According to the compliance rules, the weights of all current environmental information, all demand information, and all monitoring results are updated in real time, and the voxels corresponding to the geographic information, functional information, and real-time monitoring data at each voxel encoding are updated according to all the weights. Based on all updated voxels, generate a construction plan for the target area.

7. A spatial design optimization system based on three-dimensional voxelization, as described in any one of claims 1-6, characterized in that, The spatial design optimization system based on three-dimensional voxelization includes: The three-dimensional voxelization module is used to obtain the architectural planning scheme of the target area, determine multiple geodetic coordinates in the geodetic coordinate system based on the architectural planning scheme, and map all the geodetic coordinates to the three-dimensional voxel coordinate system to obtain multiple three-dimensional coordinates of the target area. The annotation module is used to convert all the three-dimensional coordinates into corresponding voxel grids to obtain the voxel code of each three-dimensional coordinate in the voxel grid; The weight assignment module is used to obtain various rule information of the target region, construct voxels corresponding to all data information in each rule information on each voxel encoding, and assign weights to each data information according to the defined compliance rules. A multi-source data fusion module is used to determine multiple voxel densities of all voxel codes based on the weights of each type of data information, and to predict the detection results of potential problems at the corresponding voxel codes based on all voxel densities. The scheme update module is used to adjust the building planning scheme in real time based on all the detection results, and to update the weight of all data information in all the rule information in real time to generate a construction scheme for the target area.

8. A terminal, characterized in that, The terminal includes: a memory, a processor, and a three-dimensional voxel-based spatial design optimization program stored in the memory and executable on the processor. When the three-dimensional voxel-based spatial design optimization program is executed by the processor, it implements the steps of the three-dimensional voxel-based spatial design optimization method as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a spatial design optimization program based on three-dimensional voxelization, which, when executed by a processor, implements the steps of the spatial design optimization method based on three-dimensional voxelization as described in any one of claims 1-6.

Images