Method, device and medium for identifying a thermal exposure field based on three-dimensional morphological constraints

Abstract

The application discloses a kind of based on three-dimensional form constraint's thermal exposure field identification method, device and medium, the method includes the pre-processing of target city's multi-source basic data, according to the grid unit corresponding to the target city of processed multi-source basic data is constructed;According to the digital surface model of fusion topography, building and vegetation, the sky view factor of pedestrian height and three-dimensional shelter angle are calculated per pixel, to construct three-dimensional radiation geometric relationship;According to grid unit and three-dimensional radiation geometric relationship generates three-dimensional city form constraint field;According to three-dimensional city form constraint field and the solar position parameter of current time generates pedestrian layer thermal exposure field.Based on grid unit, rely on the form constraint field generated by the digital surface model of fusion topography, building and vegetation, combine solar position parameter to reconstruct pedestrian layer thermal exposure field, consider the radiation interaction of building, vegetation, topography, can solve the defect that two-dimensional remote sensing evaluation microcosmic thermal heterogeneity is lost, and with human thermal perception deviation is big.

Description

Technical Field

[0001] This invention relates to the field of thermal exposure field identification technology, and in particular to a method, device and medium for thermal exposure field identification based on three-dimensional morphological constraints. Background Technology

[0002] With global warming leading to frequent extreme heat events in cities, significant spatial differences in urban thermal exposure exist. The street-valley geometry of high-density urban areas causes drastic microclimate differentiation at extremely small scales, making it difficult for traditional large-scale urban-rural comparative studies to capture fine-grained thermal heterogeneity within urban blocks. Currently, thermal exposure assessment methods based on two-dimensional remote sensing LST inversion utilize satellite inversion of land surface temperature, offering wide coverage and continuous temporal data, and are often used to analyze the two-dimensional correlation between the underlying surface and the thermal environment; however, the resolution is only 30m-1km, unsuitable for the street block scale; and it only reflects surface radiation temperature, ignoring building shadows, facade reflections, and the vertical structure of vegetation, thus failing to represent the true thermal perception of pedestrians.

[0003] Therefore, relying on two-dimensional remote sensing of land surface temperature (LST) has a coarse resolution, which easily homogenizes the hot and cold patches within the block and loses the microscopic thermal heterogeneity; it only characterizes the surface radiation temperature and does not take into account microscale factors such as three-dimensional shading of buildings, multiple reflections of facades, and vertical canopy transmission of vegetation, resulting in a large deviation from the actual heat load of pedestrians. Summary of the Invention

[0004] This invention provides a method, device, and medium for identifying thermal exposure fields based on three-dimensional morphological constraints. Based on multiple types of grid cells, it constructs three-dimensional radiation geometry relationships and generates a three-dimensional urban morphological constraint field by relying on a digital surface model that integrates terrain, buildings, and vegetation. It reconstructs the thermal exposure field of the pedestrian layer by combining solar position parameters, fully considering the three-dimensional radiation interaction of buildings, vegetation, and terrain. This invention can solve the defects of two-dimensional remote sensing assessment, such as loss of microscopic thermal heterogeneity and large deviation from human thermal perception.

[0005] To achieve the above objectives, embodiments of the present invention provide a method for identifying thermal exposure fields based on three-dimensional morphological constraints, including: Obtain multi-source basic data of the target city, preprocess the multi-source basic data, and construct the raster unit corresponding to the target city based on the processed multi-source basic data; Based on the digital surface model that integrates terrain, buildings and vegetation in the grid cell, a ray tracing algorithm is used to calculate the sky view factor and three-dimensional occlusion angle of pedestrian height pixel by pixel in order to construct the three-dimensional radial geometry of the target city. A three-dimensional urban morphology constraint field for the target city is generated based on the grid cells and the three-dimensional radial geometry. The pedestrian layer thermal exposure field of the target city is generated based on the three-dimensional urban morphology constraint field and the solar position parameters at the current moment.

[0006] As an improvement to the above scheme, after generating the pedestrian layer thermal exposure field of the target city, the method further includes: Based on the decoupled calculation of the pedestrian layer thermal exposure field, the rigid shadow field of buildings and the semi-transparent shadow field of vegetation in the target city are calculated, and the radiative cooling difference corresponding to the rigid shadow field of buildings and the semi-transparent shadow field of vegetation is statistically analyzed to obtain the shading and cooling data of the target city. A comprehensive heat load map of the target city is constructed based on the population density data of the target city, the heat exposure field of the pedestrian layer, and the shading and cooling data. Thermal regulation is performed on the high heat load areas of the target city based on the comprehensive heat load map and the preset conditional diffusion model.

[0007] As an improvement to the above scheme, the grid unit includes a weather-driven unit, a building rigid shielding unit, a vegetation semi-transparent shielding unit, a terrain correction unit, and a land cover thermal property unit. The process of acquiring multi-source basic data for the target city, preprocessing the multi-source basic data, and constructing raster cells corresponding to the target city based on the processed multi-source basic data includes: Acquire multi-source basic data of the target city; wherein, the multi-source basic data includes meteorological driving data, building spatial structure data, vegetation vertical structure data, topographic data and land cover data of the target city; The multi-source basic data are unified to the same coordinate system and initialized as a 1m×1m spatial grid; bilinear interpolation is used to resample and match the multi-source basic data with different resolutions in the same coordinate system to the spatial grid. The resampled meteorological driving data is mapped to the entire spatial grid to generate meteorological driving units for the target city. Based on the resampled building space structure data, assign the corresponding building height to all pixels inside the building outline, and uniformly set the transmittance of the building area to 0 to generate the building rigid shading unit of the target city. Based on the resampled vegetation vertical structure data, vegetation pixels with canopy height greater than pedestrian layer in the canopy height model are selected, the corresponding canopy height is assigned to the vegetation pixels, and the canopy transmittance is set to generate the vegetation semi-transparent shading unit of the target city. Based on the resampled terrain data, the corresponding elevation, slope and aspect are assigned to each pixel to generate terrain correction units for the target city; Based on the resampled land cover data, the corresponding shortwave albedo and longwave emissivity are assigned according to the underlying surface type of each pixel to generate the land cover thermal attribute unit of the target city.

[0008] As an improvement to the above scheme, the step of generating the pedestrian layer thermal exposure field of the target city based on the three-dimensional urban morphological constraint field and the solar position parameters at the current moment includes: Obtain the solar position parameters at the current moment; wherein, the solar position parameters include the solar altitude angle and solar azimuth angle at the current moment; Based on the three-dimensional radiation geometry of the three-dimensional urban morphology constraint field, the solar altitude angle and solar azimuth angle at the current moment, the shortwave radiation propagation process and longwave radiation exchange process are calculated pixel by pixel. The pedestrian layer thermal exposure field of the target city is generated based on the shortwave radiation propagation process and the longwave radiation exchange process.

[0009] As an improvement to the above scheme, the step of calculating the shortwave radiation propagation process and longwave radiation exchange process pixel by pixel based on the three-dimensional radiation geometry of the three-dimensional urban morphology constraint field, the solar altitude angle and solar azimuth angle at the current moment includes: Based on the current solar altitude angle and solar azimuth angle, and the three-dimensional radiation geometry of buildings, vegetation and terrain in the three-dimensional urban morphology constraint field, the shortwave radiation propagation process in the pedestrian layer space is calculated; wherein, the shortwave radiation propagation process includes direct solar radiation, sky scattered radiation, building and surface reflected radiation, and weakened radiation after transmission through the vegetation canopy. Based on the spatial relationships between building facades, ground surface, vegetation canopy, and sky view in the three-dimensional urban morphology constraint field, the longwave radiation exchange process of the pedestrian layer is calculated; wherein, the longwave radiation exchange process includes ground surface longwave radiation, building facade longwave radiation, sky longwave radiation, and vegetation canopy longwave radiation.

[0010] As an improvement to the above scheme, the step of constructing a comprehensive heat load map of the target city based on the population density data, the pedestrian layer heat exposure field, and the shading and cooling data includes: Obtain population density data for the target city, interpolate the population density data to a 1m×1m spatial grid, and generate a social sensitivity matrix for the target city. Based on the shading and cooling data, obtain shading and cooling data for different shading types and the maximum cooling data across the entire region to calculate the fitness weights for different shading types; then, superimpose the fitness weights pixel by pixel to generate the structural fitness matrix of the target city. The pedestrian layer thermal exposure field, the social sensitivity matrix, and the structural fitness matrix are normalized; a comprehensive heat load map of the target city is generated based on the normalized pedestrian layer thermal exposure field, social sensitivity matrix, and structural fitness matrix.

[0011] As an improvement to the above scheme, the step of thermal regulation of the high heat load area of ​​the target city based on the comprehensive heat load map and the preset conditional diffusion model includes: Based on the comprehensive heat load map, areas whose heat load values ​​exceed the preset heat load threshold are selected as high heat load target areas; Structural equation modeling and interpretable machine learning algorithms were used to analyze the thermal environment driving mechanism of the high heat load target area in order to determine the thermal environment effect thresholds of building density, building height and tree coverage. Extract the fixed building outlines and road red lines within the high heat load area and convert them into an immutable spatial mask matrix; The thermal environment effect threshold and the spatial mask matrix are input into a preset conditional diffusion model to generate an optimized tree layout map; The vegetation layout in the high heat load area is optimized based on the tree optimization layout map to achieve thermal regulation in the high heat load area.

[0012] As an improvement to the above scheme, the method for obtaining the preset conditional diffusion model includes: Construct a conditional diffusion model with ControlNet as the core control unit; Multiple sets of historical pedestrian thermal exposure fields of the target city were collected as sample datasets, which were divided into cold spot sample datasets and hot spot sample datasets. Using the building and tree canopy spatial distribution grid in the cold spot sample dataset as training input, and building and road boundary masks as additional conditions of ControlNet, and the corresponding street cooling efficiency labels as supervision data, the conditional diffusion model is iteratively trained to obtain the preset conditional diffusion model.

[0013] To achieve the above objectives, embodiments of the present invention provide a thermal exposure field identification device based on three-dimensional morphological constraints, comprising: The basic data acquisition module is used to acquire multi-source basic data of the target city, preprocess the multi-source basic data, and construct the raster unit corresponding to the target city based on the processed multi-source basic data. The radiation relationship construction module is used to calculate the sky view factor and three-dimensional occlusion angle of pedestrian height pixel by pixel using a ray tracing algorithm based on the digital surface model of fused terrain, buildings and vegetation in the grid cell, so as to construct the three-dimensional radiation geometry of the target city. The morphological constraint generation module is used to generate a three-dimensional urban morphological constraint field for the target city based on the grid cells and the three-dimensional radial geometric relationship. The thermal exposure field generation module is used to generate the pedestrian layer thermal exposure field of the target city based on the three-dimensional urban morphology constraint field and the solar position parameters at the current moment.

[0014] To achieve the above objectives, embodiments of the present invention also provide a computer-readable storage medium, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the above-described thermal exposure field identification method based on three-dimensional morphological constraints.

[0015] Compared with existing technologies, the present invention discloses a method, apparatus, and medium for identifying thermal exposure fields based on three-dimensional morphological constraints. This method acquires multi-source basic data of a target city, preprocesses the multi-source basic data, and constructs a grid cell corresponding to the target city based on the processed multi-source basic data. Based on the digital surface model of the grid cell, which integrates terrain, buildings, and vegetation, a ray tracing algorithm is used to calculate the sky view factor and three-dimensional occlusion angle of pedestrian height pixel by pixel to construct the three-dimensional radial geometry of the target city. A three-dimensional urban morphological constraint field is generated based on the grid cell and the three-dimensional radial geometry. Finally, a pedestrian layer thermal exposure field is generated based on the three-dimensional urban morphological constraint field and the current solar position parameters. By constructing multiple types of grid units, relying on a digital surface model that integrates terrain, buildings, and vegetation to build three-dimensional radiation geometry and generate a three-dimensional urban morphology constraint field, and combining solar position parameters to reconstruct the pedestrian layer thermal exposure field, it is possible to overcome the limitations of coarse resolution of two-dimensional surface temperature, accurately depict the micro-scale thermal heterogeneity of blocks, and improve the physical realism of three-dimensional environment modeling; by coupling multiple three-dimensional radiation constraints of terrain, buildings, and vegetation, it is possible to accurately restore the complete propagation path of long and short wave radiation at the height of pedestrian activity, reduce the deviation between thermal exposure simulation and real human thermal perception, and improve the accuracy of the pedestrian layer thermal exposure field. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating a thermal exposure field identification method based on three-dimensional morphological constraints provided in an embodiment of the present invention. Figure 2 This is a diagram of the shortwave radiation propagation process within the pedestrian layer space provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of a thermal exposure field identification device based on three-dimensional morphological constraints provided in an embodiment of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] It should be noted that the terms "comprising" and "specific" in this invention, and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to such process, method, product, or device.

[0019] Please see Figure 1 , Figure 1 This is a flowchart illustrating a method for identifying thermal exposure fields based on three-dimensional morphological constraints, provided in an embodiment of the present invention. The method includes: S1, acquire multi-source basic data of the target city, preprocess the multi-source basic data, and construct the raster unit corresponding to the target city based on the processed multi-source basic data; S2, based on the digital surface model that integrates terrain, buildings and vegetation in the grid cell, a ray tracing algorithm is used to calculate the sky view factor and three-dimensional occlusion angle of pedestrian height pixel by pixel in order to construct the three-dimensional radial geometry of the target city; S3, Generate a three-dimensional urban morphology constraint field for the target city based on the grid cells and the three-dimensional radial geometric relationship; S4. Generate the pedestrian layer thermal exposure field of the target city based on the three-dimensional urban morphology constraint field and the solar position parameters at the current moment.

[0020] For example, multi-source basic data of the target city is acquired, including meteorological driving data, 3D building vector data, tree canopy height model, digital elevation topography data, and land cover thermal attribute data. The multi-source basic data undergoes unified coordinate registration, 1-meter grid resampling, vector rasterization, and preprocessing with study area clipping. Based on the processed multi-source basic data, a raster unit corresponding to the target city is constructed. This raster unit includes rigid building shading units, semi-transparent vegetation shading units, terrain correction units, land cover thermal attribute units, and meteorological driving units. Ray tracing is used to solve the sky view factor and 3D radiation geometry relationship of the pedestrian layer pixel by pixel. A 3D urban morphology constraint field for the target city is generated based on the raster unit and the 3D radiation geometry relationship. The current solar altitude angle and solar azimuth angle are obtained as solar position parameters. Based on the 3D urban morphology constraint field, long-wave and short-wave radiation transmission coupling calculations are performed to generate a continuous pedestrian layer thermal exposure field with 1-meter resolution and a pedestrian height of 1.1 meters.

[0021] Furthermore, after generating the pedestrian-level thermal exposure field of the target city, the method further includes: S5, based on the decoupled calculation of the pedestrian layer thermal exposure field, calculate the building rigid shadow field and vegetation semi-transparent shadow field of the target city, and respectively calculate the radiation cooling difference between the building rigid shadow field and the vegetation semi-transparent shadow field to obtain the shading cooling data of the target city. S6. Construct a comprehensive heat load map of the target city based on the population density data of the target city, the heat exposure field of the pedestrian layer, and the shading and cooling data. S7. Based on the comprehensive heat load map and the preset conditional diffusion model, perform thermal regulation on the high heat load area of ​​the target city.

[0022] For example, a rigid building shadow field and a semi-transparent vegetation shadow field are generated based on the decoupled calculation of the pedestrian layer thermal exposure field, and the two are decoupled by differentiation. The rigid building shadow field represents the target city rigidly shaded by buildings, and the semi-transparent vegetation shadow field represents the target city semi-transparently shaded by tree canopy. The coverage area, duration, radiation reduction intensity, and thermal exposure reduction of the two types of shadows are calculated respectively. The difference in radiative cooling corresponding to the rigid building shadow field and the semi-transparent vegetation shadow field is statistically analyzed to obtain the shading and cooling data of the target city, thereby identifying the contribution of different shading mechanisms to the improvement of the local thermal environment.

[0023] Specifically, the grid unit includes a weather-driven unit, a building rigid shading unit, a vegetation semi-transparent shading unit, a terrain correction unit, and a land cover thermal property unit; Step S1 includes: S11, acquire multi-source basic data of the target city; wherein, the multi-source basic data includes meteorological driving data, building spatial structure data, vegetation vertical structure data, topographic data and land cover data of the target city; S12, unify the multi-source basic data into the same coordinate system and initialize it as a 1m×1m spatial grid; use bilinear interpolation to resample and match the multi-source basic data of different resolutions in the same coordinate system to the spatial grid; S13, map the resampled meteorological driving data to the entire spatial grid to generate meteorological driving units for the target city; S14. Based on the resampled building space structure data, assign the corresponding building height to all pixels inside the building outline, and uniformly set the transmittance of the building area to 0 to generate the building rigid shading unit of the target city. S15. Based on the resampled vegetation vertical structure data, filter the vegetation pixels in the canopy height model whose canopy height is greater than the pedestrian layer, assign the corresponding canopy height to the vegetation pixels, and set the canopy transmittance to generate the vegetation semi-transparent shading unit of the target city. S16, Based on the resampled terrain data, assign the corresponding elevation, slope and aspect to each pixel to generate the terrain correction unit for the target city; S17. Based on the resampled land cover data, assign the corresponding shortwave albedo and longwave emissivity to the underlying surface type of each pixel to generate the land cover thermal attribute unit of the target city.

[0024] For example, the macroscopic physical background of the target city is acquired by selecting reanalysis data, meteorological station observation data, or other meteorological data sources with temporal continuity and global coverage. Key parameters such as 2-meter air temperature, relative humidity, 10-meter wind speed (including U / V components), surface downdraft solar radiation, clear-sky direct solar radiation, total precipitation, and surface air pressure are then selected and extracted to provide macroscopic atmospheric constraints for radiative transfer calculations in three-dimensional space. In terms of digital representation of the city's micro-geometry, multi-dimensional physical environment data is integrated to reconstruct the true three-dimensional mechanisms of the city. This includes building spatial structure data, vegetation vertical structure data, topographic data, and land cover data. Building height is used to determine rigid shading boundaries in urban space, vegetation canopy height is used to characterize the vertical shading structure of the canopy, topographic elevation is used to correct the impact of surface undulations on radiation paths, and land cover type is used to assign thermophysical properties to different underlying surfaces. After acquiring multi-source data, a unified spatial topology grid is established. This grid does not simply unify data of different resolutions to the same scale through interpolation. Instead, it establishes raster cells for radiative transfer calculations based on the spatial constraints between building boundaries, vegetation canopy boundaries, topographic relief, and meteorological driving fields. Based on these raster cells, a three-dimensional radiative geometric relationship is constructed between buildings, vegetation, topography, and the sky view to build a three-dimensional urban constraint field for pedestrian thermal exposure calculations. This three-dimensional radiative geometric relationship includes the direct solar path, inter-building shading relationships, canopy transmission relationships, topographic shading relationships, and sky view attenuation relationships.

[0025] It is worth noting that the three-dimensional radiation geometry field is used to describe the propagation boundaries of shortwave and longwave radiation under complex urban morphological conditions. This three-dimensional urban constraint field includes rigid building shielding units, semi-transparent vegetation shielding units, terrain correction units, land cover thermal attribute units, and meteorological driving units, which can provide a unified spatial basis for subsequent reconstruction of continuous thermal exposure fields.

[0026] More specifically, the grid unit includes a weather-driven unit, a building rigid shading unit, a vegetation semi-transparent shading unit, a terrain correction unit, and a land cover thermal property unit; Taking a selected high-density block within a target city (e.g., a spatial range of 1000m × 1000m) as an example, a 1m × 1m two-dimensional topological grid is initialized as a unified spatial calculation benchmark. When constructing terrain correction units and land cover thermal attribute units, a 5m resolution digital elevation model and 10m resolution land cover data are projected onto this benchmark grid. A bilinear interpolation algorithm is used to assign each 1m pixel an elevation value, slope, aspect, and shortwave albedo and longwave emissivity parameters of a specific underlying surface (e.g., asphalt, grassland), thereby defining the basic thermophysical boundary of the land surface. Building vector data with height attributes are rasterized and superimposed onto the benchmark grid. For pixels falling within the building outline, an absolute building height value is assigned, and their transmittance is set to zero, thus constructing a rigid building shading unit that forms a rigid physical shading boundary. A 1m resolution tree canopy height model is introduced to extract vegetation pixels above the pedestrian level (e.g., set to 1.1m). Unlike the absolute shading of buildings, these vegetation pixels are simultaneously assigned canopy height values ​​and specific shortwave radiation transmittance parameters (e.g., setting the canopy transmittance to 0.05 under a specific canopy closure), thus forming vegetation semi-transparent shading units with semi-transmissive physical properties. After meshing, based on a comprehensive digital surface model including terrain, buildings, and trees, ray tracing algorithms are used to calculate the sky view factor and three-dimensional shading angle at pedestrian level pixel by pixel, defining the proportion of the visible sky dome for each pixel. The acquired target city-level macro-meteorological parameters are used as dynamic inputs in the time dimension and mapped to the entire spatial grid to form meteorological driven units. Thus, multi-dimensional parameters including geometric shading, medium transmission, surface thermal properties, and macro-meteorology are tightly coupled in the same high-resolution three-dimensional mesh matrix, fully constructing a three-dimensional urban constraint field. This provides a unified and physically meaningful spatial basis for subsequent accurate calculation of the physical fields of direct solar path, multiple reflections, and continuous thermal exposure.

[0027] Specifically, step S4 includes: S41, Obtain the solar position parameters at the current moment; wherein, the solar position parameters include the solar altitude angle and solar azimuth angle at the current moment; S42, based on the three-dimensional radiation geometry of the three-dimensional urban morphology constraint field, the solar altitude angle and solar azimuth angle at the current moment, calculate the shortwave radiation propagation process and longwave radiation exchange process pixel by pixel; S43, Generate the pedestrian layer thermal exposure field of the target city based on the shortwave radiation propagation process and the longwave radiation exchange process.

[0028] For example, based on the solar position parameters at the current moment, the solar altitude angle and solar azimuth angle are determined. Combined with the three-dimensional radiation geometry field formed by buildings, vegetation, and terrain, the shortwave radiation propagation process within the pedestrian layer is calculated. Based on the spatial relationships between building facades, the ground surface, vegetation canopy, and the sky, the longwave radiation exchange process within the pedestrian layer is calculated. Based on the above calculation results of shortwave and longwave radiation, a pedestrian layer thermal exposure field is generated to characterize the actual radiative thermal exposure experienced by the human body in urban microspaces. This thermal exposure field is expressed in the form of a high-resolution continuous spatial grid, capable of reflecting microscale thermal exposure differences in building edges, street valleys, under tree canopies, open plazas, and road spaces.

[0029] More specifically, step S42 includes: S421, Based on the solar altitude angle and solar azimuth angle at the current moment, and the three-dimensional radiation geometry of buildings, vegetation and terrain in the three-dimensional urban morphology constraint field, calculate the shortwave radiation propagation process in the pedestrian layer space; wherein, the shortwave radiation propagation process includes direct solar radiation, sky scattered radiation, building and surface reflected radiation, and weakened radiation after transmission through the vegetation canopy. S422, Based on the spatial relationship between building facades, ground surface, vegetation canopy and sky view in the three-dimensional urban morphology constraint field, calculate the long-wave radiation exchange process of the pedestrian layer; wherein, the long-wave radiation exchange process includes ground surface long-wave radiation, building facade long-wave radiation, sky long-wave radiation and vegetation canopy long-wave radiation.

[0030] For example, such as Figure 2 As shown, Figure 2 This invention provides a diagram illustrating the shortwave radiation propagation process within the pedestrian zone. Based on the current solar position parameters, the solar altitude angle and solar azimuth angle are determined. Combined with the three-dimensional radiation geometry field formed by buildings, vegetation, and terrain, the shortwave radiation propagation process within the pedestrian zone is calculated. This shortwave radiation propagation process includes direct solar radiation, sky-scattered radiation, radiation reflected from buildings and the ground surface, and attenuated radiation transmitted through the vegetation canopy. Simultaneously, based on the spatial relationships between building facades, the ground surface, the vegetation canopy, and the sky, the longwave radiation exchange process within the pedestrian zone is calculated. This longwave radiation exchange process includes longwave radiation from the ground surface, building facades, the sky, and the vegetation canopy.

[0031] Specifically, step S6 includes: S61, Obtain population density data of the target city, interpolate the population density data to a 1m×1m spatial grid, and generate the social sensitivity matrix of the target city; S62, based on the shading and cooling data, obtain shading and cooling data for different shading types and the maximum cooling data for the entire region, and calculate the fitness weights for different shading types; superimpose the fitness weights pixel by pixel to generate the structural fitness matrix of the target city; S63, normalize the pedestrian layer thermal exposure field, the social sensitivity matrix, and the structural fitness matrix; generate a comprehensive heat load map of the target city based on the normalized pedestrian layer thermal exposure field, social sensitivity matrix, and structural fitness matrix.

[0032] For example, after reconstructing the pedestrian layer's thermal exposure field, a spatial coupling algorithm is used to deeply couple the microscopic physical thermal field with urban socio-demographic attributes and the resilience of spatial structures, thereby solving for a high-precision socio-physical thermal load continuum field. This approach aims to overcome the limitations of traditional two-dimensional physical meteorological indicators in characterizing complex urban environments. Its core lies in spatially mapping and logically fusing the fine physical characteristics captured by the three-dimensional radiation field with the social attributes and morphological features within the city, thereby comprehensively, qualitatively, and quantitatively assessing the overall thermal load pressure faced by the population within the urban micro-space. A comprehensive diagnostic model encompassing physical thermal exposure components, social sensitivity components, and structural adaptability components needs to be constructed to achieve precise localization of urban micro-environment thermal load risks.

[0033] In the specific quantitative definition of the indicators, the physical heat exposure component is derived from the results of three-dimensional radiative transfer simulation. It deeply integrates direct solar radiation, sky diffuse radiation, and the comprehensive field obtained from multiple reflections and long-wave radiation exchanges generated by building facades and vegetation canopies. This accurately recreates the instantaneous radiative heat exposure felt by pedestrians in complex three-dimensional street canyons. The social sensitivity component incorporates sociological considerations and is characterized using gridded high-precision population density data. It aims to reflect the spatial distribution base of the potential population affected by high-temperature disasters within a specific spatial unit under extreme heat stress conditions. The structural fitness component is defined as the potential field of microclimate buffering provided by the urban three-dimensional built environment entities. It is quantified by identifying the availability of building shadows and tree shading resources in three-dimensional space and combining the empirical weights of different shading types on the reduction intensity of the three-dimensional radiation field. This calculation method does not rely on the instantaneous cooling value at a certain moment, but focuses on reflecting the structural supply capacity of the urban three-dimensional spatial texture in providing thermal protection.

[0034] To differentiate the heat mitigation effectiveness of different shading types, data-driven differential weights are assigned to different types of shading in the space based on their cooling values. These weights represent the relative contribution of each shading form to microclimate regulation and heat mitigation capabilities. Specifically, the formula for calculating the weight of a particular shading type is as follows: , In the formula, This represents the fitness weight corresponding to a certain type of shading; This represents the difference in radiative cooling produced by a single type of shading. This represents the maximum total temperature reduction that can be achieved by combining all shading types across the entire area.

[0035] After quantifying and weighting the basic data of each dimension, a spatial superposition operation of the multi-source field is performed. By multiplicatively coupling physical heat exposure, social sensitivity and the "1 minus fitness" term determined by the shading structure, a comprehensive heat load map with pixel-level accuracy is finally calculated.

[0036] To illustrate the construction process of this socio-physical heat load continuum, a typical high-density urban block (e.g., with a spatial range of 1000m × 1000m) is used as an example. Regarding data acquisition and processing, a 1-meter resolution comprehensive radiation field matrix is ​​first extracted as physical heat exposure data. Simultaneously, census data of the target city is introduced, and its scale is reduced to a 1-meter resolution population density matrix using spatial area interpolation as social sensitivity data. Furthermore, combining building vectors and a 1-meter resolution tree canopy height model, rigid building shadow patches and semi-transparent tree canopy shading patches are accurately extracted. In the specific spatial calculation method, the radiation temperature difference between the target city directly exposed to sunlight and various shaded target cities in the aforementioned physical field is compared and analyzed to dynamically calculate specific cooling values ​​and allocate weights. For example, after physical model calculation, the average radiation cooling value of a specific tree canopy shading pixel in the target city is found to be 15℃, and the average cooling value of a building shadow pixel is 10℃. Simultaneously, the maximum cooling extreme value calculated by the physical field (e.g., 20℃) is extracted as a theoretical benchmark. Based on the aforementioned weighting formula, quantitative weights of 0.75 and 0.5 are precisely assigned to the tree canopy pixel and building shadow pixel, respectively, thereby constructing a structural fitness matrix that reflects the true resilience of the space. Subsequently, the three matrices are uniformly normalized to a dimensionless range of 0 to 1, and a pixel-by-pixel spatial multiplicative aggregation operation is performed: the normalized heat exposure value of each pixel is multiplied by the sensitivity value to obtain the initial heat stress field strength, and then multiplied by the spatial negative attenuation operator composed of "1 minus fitness weight" to complete the matrix cancellation calculation. This embodiment of the invention effectively corrects the technical bias of traditional two-dimensional meteorological assessment, which is prone to overestimating the heat load of high-density blocks with good spatial shading, by introducing a quantitative mechanism for the physical stripping and blocking of local radiation energy by three-dimensional built entities. At the pixel level, it accurately screens out real vulnerable spaces with high physical radiation, dense population, and extreme lack of shading resources. The final output of the social-physical heat load continuous map anchors precise physical intervention targets for subsequent intelligent spatial optimization solutions.

[0037] Specifically, step S7 includes: S71, based on the comprehensive heat load map, select areas whose heat load values ​​exceed the preset heat load threshold as high heat load target areas; S72, using structural equation modeling and interpretable machine learning algorithms to analyze the thermal environment driving mechanism of the high heat load target area, in order to determine the thermal environment effect thresholds of building density, building height and tree coverage; S73, extract the fixed building outlines and road red lines within the high heat load area and convert them into an immutable spatial mask matrix; S74, input the thermal environment effect threshold and the spatial mask matrix into a preset conditional diffusion model to generate a tree optimization layout map; S75, optimize the vegetation layout of the high heat load area according to the tree optimization layout map to achieve thermal regulation of the high heat load area.

[0038] For example, in high heat load areas, spatial regulation schemes are solved while maintaining existing urban morphological constraints through driving mechanism analysis and artificial intelligence generation techniques. For instance, in the driving mechanism analysis stage, structural equation modeling (SEM) and SHAP (SHapley Additive exPlanations) interpretable machine learning algorithms are coupled to construct a collaborative analysis framework that complements linear causality and nonlinear features. The urban micro-thermal environment is profoundly influenced by building morphology and vegetation elements, and there are complex interactions between the two. Utilizing the path analysis capabilities of SEM, the global causal transmission paths of the influence of tree distribution characteristics and building three-dimensional morphology on the microclimate thermal environment are identified and verified at the macro level. Furthermore, a tree-based machine learning algorithm and a SHAP interpreter are introduced to deeply explore nonlinear response relationships at the microscale. By extracting the thermal environment effect thresholds of key morphological variables such as building density, building height, and tree coverage, we can deeply reveal the dynamic game mechanism between the shadow protection effect (i.e., effectively blocking downward solar shortwave radiation) and the radiation trap effect (i.e., hindering the dissipation of ground longwave radiation and the circulation of breeze) generated by buildings in high-density urban areas. This will establish the optimization boundary and design rules in both physical and statistical sense for subsequent intelligent space regulation.

[0039] After clarifying the driving mechanism and threshold boundary of microclimate heat exposure, a conditional diffusion model from generative artificial intelligence is used to generate schemes. Based on the refined classification results of pedestrian layer heat exposure fields, specific blocks with excellent thermal comfort are extracted as positive sample source domains from a vast urban sample space. Through deep learning of these source domain data, the model can extract the spatial topological distribution patterns of building vegetation that generate excellent microclimates. A conditional diffusion model with ControlNet as the core control unit is constructed to perform intelligent spatial inference. Considering the practical dilemmas faced by existing urban renewal, the model transforms the rigid spatial entities such as the unalterable existing building base outlines and road red line networks in high heat load areas into spatial mask matrices, which are directly injected into the ControlNet network layer as hard constraints. In the reverse denoising generation process of the diffusion model, this conditional injection mechanism forces the neural network to transfer the high-quality greening configuration strategies learned in the source domain across target cities while strictly avoiding the above-mentioned physical hard constraints. It can automatically generate a refined tree layout design map with thermal regulation efficiency and in line with realistic planning constraints on limited permeable open spaces in high heat load areas, realizing the transformation from micro-heat load diagnosis to spatial renewal blueprint.

[0040] Specifically, the method for obtaining the preset conditional diffusion model includes: Construct a conditional diffusion model with ControlNet as the core control unit; Multiple sets of historical pedestrian thermal exposure fields of the target city were collected as sample datasets, which were divided into cold spot sample datasets and hot spot sample datasets. Using the building and tree canopy spatial distribution grid in the cold spot sample dataset as training input, and building and road boundary masks as additional conditions of ControlNet, and the corresponding street cooling efficiency labels as supervision data, the conditional diffusion model is iteratively trained to obtain the preset conditional diffusion model.

[0041] For example, after obtaining the thermal exposure field of the human body, a method for dividing the thermal exposure field into cold and hot spot regions can be used. For instance, local spatial autocorrelation analysis methods (such as Getis-Ord) can be employed. The system uses statistical measures to identify statistically significant heat exposure anomalies by calculating a spatial weight matrix. It calculates the local sum of a location's attribute value and its surrounding neighbor's attribute values, comparing this sum with the global sum, and outputs the corresponding Z-score (standard normal score) and P-value (significance probability). Based on set confidence intervals (e.g., 90%, 95%, 99%), the continuous human-layer heat exposure field is divided into significant hotspots (HS), significant coldspots (CS), and insignificant areas, thus identifying macroscopic radiative energy accumulation patches. Local physical field features are extracted and a microscopic classification system is constructed. To capture the complex microscale heterogeneity within the block, the identified significant hotspots and coldspots are used as basic analysis units. Three core physical field discrimination indicators are extracted within each unit: average temperature, temperature standard deviation, and the area proportion of extreme heat exposure pixels. The average temperature reflects the overall heat exposure baseline energy level of the unit; the standard deviation characterizes the dispersion and spatial uniformity of the internal radiation distribution; and the area proportion quantifies the spatial coverage intensity of extreme physical features (extreme high or low temperatures).

[0042] Objective classification thresholds were established by calculating the quartiles (such as the median Q2, upper quartile Q3, etc.) of the three indicators mentioned above in all significant cold and hot spot grids, respectively, thus constructing a four-quadrant discrimination rule. This rule divides macroscopic cold and hot patches into microscopic types with clear spatial morphology and intervention guidance significance. The classification results directly reflect the defects in urban physical morphology and can be used to support the targeted identification of key target cities and the generation of intelligent microclimate regulation schemes.

[0043] This invention discloses a method for identifying thermal exposure fields based on three-dimensional morphological constraints. The method involves acquiring multi-source basic data of a target city, preprocessing the data, and constructing a grid cell corresponding to the target city based on the processed data. Using a digital surface model that integrates terrain, buildings, and vegetation within the grid cell, a ray tracing algorithm is employed to calculate the sky view factor and three-dimensional occlusion angle of pedestrian height pixel by pixel, thereby constructing the three-dimensional radial geometry of the target city. A three-dimensional urban morphological constraint field is generated based on the grid cell and the three-dimensional radial geometry. Finally, a pedestrian-layer thermal exposure field is generated based on the three-dimensional urban morphological constraint field and the current solar position parameters. By constructing multiple types of grid units, relying on a digital surface model that integrates terrain, buildings, and vegetation to build three-dimensional radiation geometry and generate a three-dimensional urban morphology constraint field, and combining solar position parameters to reconstruct the pedestrian layer thermal exposure field, it is possible to overcome the limitations of coarse resolution of two-dimensional surface temperature, accurately depict the micro-scale thermal heterogeneity of blocks, and improve the physical realism of three-dimensional environment modeling; by coupling multiple three-dimensional radiation constraints of terrain, buildings, and vegetation, it is possible to accurately restore the complete propagation path of long and short wave radiation at the height of pedestrian activity, reduce the deviation between thermal exposure simulation and real human thermal perception, and improve the accuracy of the pedestrian layer thermal exposure field.

[0044] See Figure 3 , Figure 3 This is a schematic diagram of a thermal exposure field identification device 10 based on three-dimensional morphological constraints provided in an embodiment of the present invention. The thermal exposure field identification device 10 based on three-dimensional morphological constraints includes: The basic data acquisition module 11 is used to acquire multi-source basic data of the target city, preprocess the multi-source basic data, and construct the raster unit corresponding to the target city based on the processed multi-source basic data. The radiation relationship construction module 12 is used to calculate the sky view factor and three-dimensional occlusion angle of pedestrian height per pixel using a ray tracing algorithm based on the digital surface model of fused terrain, buildings and vegetation in the grid cell, so as to construct the three-dimensional radiation geometry of the target city. The morphological constraint generation module 13 is used to generate a three-dimensional urban morphological constraint field for the target city based on the grid unit and the three-dimensional radial geometric relationship. The thermal exposure field generation module 14 is used to generate the pedestrian layer thermal exposure field of the target city based on the three-dimensional urban morphology constraint field and the solar position parameters at the current moment.

[0045] The thermal exposure field identification device 10 based on three-dimensional morphological constraints provided in this embodiment of the invention can realize all the processes of the thermal exposure field identification method based on three-dimensional morphological constraints in the above embodiments. The functions and technical effects of each module in the device are the same as those of the thermal exposure field identification method based on three-dimensional morphological constraints in the above embodiments, and will not be repeated here.

[0046] This invention also provides a computer-readable storage medium, which includes a stored computer program, wherein the computer program, when running, controls the device where the computer-readable storage medium is located to execute the thermal exposure field identification method based on three-dimensional morphological constraints as described in the above embodiments.

[0047] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for identifying thermal exposure fields based on three-dimensional morphological constraints, characterized in that, include: Obtain multi-source basic data of the target city, preprocess the multi-source basic data, and construct the raster unit corresponding to the target city based on the processed multi-source basic data; Based on the digital surface model that integrates terrain, buildings and vegetation in the grid cell, a ray tracing algorithm is used to calculate the sky view factor and three-dimensional occlusion angle of pedestrian height pixel by pixel in order to construct the three-dimensional radial geometry of the target city. A three-dimensional urban morphology constraint field for the target city is generated based on the grid cells and the three-dimensional radial geometry. The pedestrian layer thermal exposure field of the target city is generated based on the three-dimensional urban morphology constraint field and the solar position parameters at the current moment; Based on the decoupled calculation of the pedestrian layer thermal exposure field, the rigid shadow field of buildings and the semi-transparent shadow field of vegetation in the target city are calculated, and the radiative cooling difference corresponding to the rigid shadow field of buildings and the semi-transparent shadow field of vegetation is statistically analyzed to obtain the shading and cooling data of the target city. A comprehensive heat load map of the target city is constructed based on the population density data of the target city, the heat exposure field of the pedestrian layer, and the shading and cooling data. Thermal regulation is performed on the high heat load area of ​​the target city based on the comprehensive heat load map and the preset conditional diffusion model; The step of constructing a comprehensive heat load map of the target city based on the population density data, the pedestrian layer heat exposure field, and the shading and cooling data includes: Obtain population density data for the target city, interpolate the population density data to a 1m×1m spatial grid, and generate a social sensitivity matrix for the target city. Based on the shading and cooling data, obtain shading and cooling data for different shading types and the maximum cooling data across the entire region to calculate the fitness weights for different shading types; then, superimpose the fitness weights pixel by pixel to generate the structural fitness matrix of the target city. The pedestrian layer thermal exposure field, the social sensitivity matrix, and the structural fitness matrix are normalized; a comprehensive heat load map of the target city is generated based on the normalized pedestrian layer thermal exposure field, social sensitivity matrix, and structural fitness matrix. The step of thermal regulation of high-heat-load areas in the target city based on the comprehensive heat load map and a preset conditional diffusion model includes: Based on the comprehensive heat load map, areas whose heat load values ​​exceed the preset heat load threshold are selected as high heat load target areas; Structural equation modeling and interpretable machine learning algorithms were used to analyze the thermal environment driving mechanism of the high heat load target area in order to determine the thermal environment effect thresholds of building density, building height and tree coverage. Extract the fixed building outlines and road red lines within the high heat load area and convert them into an immutable spatial mask matrix; The thermal environment effect threshold and the spatial mask matrix are input into a preset conditional diffusion model to generate an optimized tree layout map; Optimize the vegetation layout of the high heat load area according to the tree optimization layout map to achieve thermal regulation of the high heat load area; The method for obtaining the preset conditional diffusion model includes: Construct a conditional diffusion model with ControlNet as the core control unit; Multiple sets of historical pedestrian thermal exposure fields of the target city were collected as sample datasets, which were divided into cold spot sample datasets and hot spot sample datasets. Using the building and tree canopy spatial distribution grid in the cold spot sample dataset as training input, and building and road boundary masks as additional conditions of ControlNet, and the corresponding street cooling efficiency labels as supervision data, the conditional diffusion model is iteratively trained to obtain the preset conditional diffusion model.

2. The thermal exposure field identification method based on three-dimensional morphological constraints as described in claim 1, characterized in that, The grid unit includes a weather-driven unit, a building rigid shielding unit, a vegetation semi-transparent shielding unit, a terrain correction unit, and a land cover thermal property unit. The process of acquiring multi-source basic data for the target city, preprocessing the multi-source basic data, and constructing raster cells corresponding to the target city based on the processed multi-source basic data includes: Acquire multi-source basic data of the target city; wherein, the multi-source basic data includes meteorological driving data, building spatial structure data, vegetation vertical structure data, topographic data and land cover data of the target city; The multi-source basic data are unified to the same coordinate system and initialized as a 1m×1m spatial grid; bilinear interpolation is used to resample and match the multi-source basic data with different resolutions in the same coordinate system to the spatial grid. The resampled meteorological driving data is mapped to the entire spatial grid to generate meteorological driving units for the target city. Based on the resampled building space structure data, assign the corresponding building height to all pixels inside the building outline, and uniformly set the transmittance of the building area to 0 to generate the building rigid shading unit of the target city. Based on the resampled vegetation vertical structure data, vegetation pixels with canopy height greater than pedestrian layer in the canopy height model are selected, the corresponding canopy height is assigned to the vegetation pixels, and the canopy transmittance is set to generate the vegetation semi-transparent shading unit of the target city. Based on the resampled terrain data, the corresponding elevation, slope and aspect are assigned to each pixel to generate terrain correction units for the target city; Based on the resampled land cover data, the corresponding shortwave albedo and longwave emissivity are assigned according to the underlying surface type of each pixel to generate the land cover thermal attribute unit of the target city.

3. The thermal exposure field identification method based on three-dimensional morphological constraints as described in claim 1, characterized in that, The step of generating the pedestrian layer thermal exposure field of the target city based on the three-dimensional urban morphology constraint field and the solar position parameters at the current moment includes: Obtain the solar position parameters at the current moment; wherein, the solar position parameters include the solar altitude angle and solar azimuth angle at the current moment; Based on the three-dimensional radiation geometry of the three-dimensional urban morphology constraint field, the solar altitude angle and solar azimuth angle at the current moment, the shortwave radiation propagation process and longwave radiation exchange process are calculated pixel by pixel. The pedestrian layer thermal exposure field of the target city is generated based on the shortwave radiation propagation process and the longwave radiation exchange process.

4. The thermal exposure field identification method based on three-dimensional morphological constraints as described in claim 3, characterized in that, The step of calculating the shortwave radiation propagation process and longwave radiation exchange process pixel by pixel based on the three-dimensional radiation geometry of the three-dimensional urban morphology constraint field, the solar altitude angle and solar azimuth angle at the current moment, includes: Based on the current solar altitude angle and solar azimuth angle, and the three-dimensional radiation geometry of buildings, vegetation and terrain in the three-dimensional urban morphology constraint field, the shortwave radiation propagation process in the pedestrian layer space is calculated; wherein, the shortwave radiation propagation process includes direct solar radiation, sky scattered radiation, building and surface reflected radiation, and weakened radiation after transmission through the vegetation canopy. Based on the spatial relationships between building facades, ground surface, vegetation canopy, and sky view in the three-dimensional urban morphology constraint field, the longwave radiation exchange process of the pedestrian layer is calculated; wherein, the longwave radiation exchange process includes ground surface longwave radiation, building facade longwave radiation, sky longwave radiation, and vegetation canopy longwave radiation.

5. A thermal exposure field identification device based on three-dimensional morphological constraints, characterized in that, include: The basic data acquisition module is used to acquire multi-source basic data of the target city, preprocess the multi-source basic data, and construct the raster unit corresponding to the target city based on the processed multi-source basic data. The radiation relationship construction module is used to calculate the sky view factor and three-dimensional occlusion angle of pedestrian height pixel by pixel using a ray tracing algorithm based on the digital surface model of fused terrain, buildings and vegetation in the grid cell, so as to construct the three-dimensional radiation geometry of the target city. The morphological constraint generation module is used to generate a three-dimensional urban morphological constraint field for the target city based on the grid cells and the three-dimensional radial geometric relationship. The thermal exposure field generation module is used to generate the pedestrian layer thermal exposure field of the target city based on the three-dimensional urban morphology constraint field and the solar position parameters at the current moment. The cooling data acquisition module is used to decouple and calculate the rigid shadow field of buildings and the semi-transparent shadow field of vegetation in the target city based on the thermal exposure field of the pedestrian layer, and to calculate the radiation cooling difference between the rigid shadow field of buildings and the semi-transparent shadow field of vegetation respectively, so as to obtain the shading cooling data of the target city. The heat load map construction module is used to construct a comprehensive heat load map of the target city based on the population density data of the target city, the pedestrian layer heat exposure field, and the shading and cooling data. The load area adjustment module is used to perform thermal adjustment on the high heat load area of ​​the target city according to the comprehensive heat load map and the preset conditional diffusion model; The heat load diagram construction module is used for: Obtain population density data for the target city, interpolate the population density data to a 1m×1m spatial grid, and generate a social sensitivity matrix for the target city. Based on the shading and cooling data, obtain shading and cooling data for different shading types and the maximum cooling data across the entire region to calculate the fitness weights for different shading types; then, superimpose the fitness weights pixel by pixel to generate the structural fitness matrix of the target city. The pedestrian layer thermal exposure field, the social sensitivity matrix, and the structural fitness matrix are normalized; a comprehensive heat load map of the target city is generated based on the normalized pedestrian layer thermal exposure field, social sensitivity matrix, and structural fitness matrix. The load area adjustment module is used for: Based on the comprehensive heat load map, areas whose heat load values ​​exceed the preset heat load threshold are selected as high heat load target areas; Structural equation modeling and interpretable machine learning algorithms were used to analyze the thermal environment driving mechanism of the high heat load target area in order to determine the thermal environment effect thresholds of building density, building height and tree coverage. Extract the fixed building outlines and road red lines within the high heat load area and convert them into an immutable spatial mask matrix; The thermal environment effect threshold and the spatial mask matrix are input into a preset conditional diffusion model to generate an optimized tree layout map; Optimize the vegetation layout of the high heat load area according to the tree optimization layout map to achieve thermal regulation of the high heat load area; The method for obtaining the preset conditional diffusion model includes: Construct a conditional diffusion model with ControlNet as the core control unit; Multiple sets of historical pedestrian thermal exposure fields of the target city were collected as sample datasets, which were divided into cold spot sample datasets and hot spot sample datasets. Using the building and tree canopy spatial distribution grid in the cold spot sample dataset as training input, and building and road boundary masks as additional conditions of ControlNet, and the corresponding street cooling efficiency labels as supervision data, the conditional diffusion model is iteratively trained to obtain the preset conditional diffusion model.

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device containing the computer-readable storage medium to perform the thermal exposure field identification method based on any one of claims 1-4.

Images