WRF and XGBoost coupling simulation-based built-up area thermal comfort fine evaluation method, system and terminal

Abstract

The invention discloses a built-up area thermal comfort fine evaluation method, system and terminal based on WRF and XGBoost coupling simulation, and the method comprises the steps: calculating urban morphological parameters based on an LCZ map through geographic space information data; improving the WRF model by using the LCZ map and the urban morphological parameters to obtain a target WRF model; respectively establishing historical microclimate data sets corresponding to the air temperature, the relative humidity and the wind speed, and training and testing the XGBoost model by using the historical microclimate data sets to obtain a target XGBoost model; inputting the simulation result to be evaluated into the target XGBoost model to obtain a target air temperature, a target relative humidity and a target wind speed; and according to the target air temperature, the target relative humidity and the target wind speed, calculating a thermal comfort evaluation index of the built-up area. According to the method, multi-element, large-range, high-precision and high-temporal-spatial-resolution built-up area thermal comfort evaluation is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of urban thermal comfort assessment, and in particular to a method, system, terminal and computer-readable storage medium for refined assessment of thermal comfort in built-up areas based on WRF (Weather Research and Forecasting Model) and XGBoost (eXtreme Gradient Boosting) coupled simulation. Background Art

[0002] Thermal comfort in built-up areas refers to the degree of comfort people experience with local climate conditions (such as temperature, humidity, wind speed, and solar radiation) in the built-up environment of the city. At present, the indicators for thermal comfort assessment in large-scale built-up areas are mainly calculated by linear equations constructed based on meteorological variables. These linear equations are mainly constructed by temperature or temperature and humidity, and lack consideration of wind speed variables. Existing studies have shown that wind speed is one of the important factors affecting thermal comfort, especially in coastal cities with frequent sea breezes. However, due to the availability of data, wind speed variables are difficult to apply to thermal comfort assessment indicators, that is, the current thermal comfort assessment in built-up areas is insufficient in terms of characterization ability. In addition, the meteorological data required for thermal comfort assessment in built-up areas are often obtained through fixed-point or mobile meteorological monitoring methods. These methods are highly dependent on equipment deployment and manual maintenance. The temporal and spatial resolution is difficult to adjust according to the needs of thermal comfort assessment, and it often costs a lot of money to collect data over a large range and for a long time. That is, the current thermal comfort assessment in built-up areas is insufficient in terms of assessment scope and precision.

[0003] Therefore, the prior art still needs to be improved and developed. Summary of the invention

[0004] The main purpose of the present invention is to provide a method, system, terminal and computer-readable storage medium for refined evaluation of thermal comfort in built-up areas based on WRF and XGBoost coupling simulation, aiming to solve the problem that existing thermal comfort evaluation methods are difficult to balance characterization capability, evaluation scope and precision.

[0005] To achieve the above-mentioned purpose of the invention, the present invention provides a refined evaluation method for thermal comfort in built-up areas based on WRF and XGBoost coupling simulation, and the refined evaluation method for thermal comfort in built-up areas based on WRF and XGBoost coupling simulation includes: Based on the local climate zoning map, the urban morphological parameters of each grid unit in the built-up area are calculated using geospatial information data; Improve the WRF model using the local climate zoning map and a plurality of the urban morphology parameters to obtain a target WRF model; Simulating weather using the target WRF model, wherein the target WRF model outputs simulation results, wherein the simulation results include historical simulation results and simulation results to be evaluated, and both the historical simulation results and the simulation results to be evaluated include simulated temperature, simulated relative humidity, and simulated wind speed; Establishing historical microclimate data sets corresponding to air temperature, relative humidity, and wind speed respectively, and using a plurality of the historical microclimate data sets to train and test an XGBoost model to obtain a target XGBoost model; Inputting the simulation result to be evaluated into the target XGBoost model, the target XGBoost model outputs a target air temperature, a target relative humidity, and a target wind speed; The thermal comfort assessment index of the built-up area is calculated based on the target air temperature, the target relative humidity and the target wind speed.

[0006] Optionally, the method of calculating the urban morphological parameters of each grid unit in the built-up area based on the local climate zoning map using geographic spatial information data specifically includes: Obtaining a LCZ map and geospatial information data, wherein the geospatial information data includes building data, land cover data, and terrain elevation data; Through a GIS software platform, the building data, the land cover data and the terrain elevation data are superimposed on the LCZ map to calculate the urban morphological parameters of each grid cell in each built-up area class; Among them, the urban morphology parameters include building coverage, building height, building height standard deviation, building width, total building volume, total building area, street width, sky visibility, permeability, water coverage and altitude.

[0007] Optionally, the improving the WRF model by using the local climate zoning map and the plurality of urban morphology parameters to obtain a target WRF model specifically includes: Set the simulation period, nested domain and projection coordinates of the WRF model; The local climate zoning map is coupled with the land cover data of the lowest nested domain of the WRF model, and the land cover classification scheme of the lowest nested domain of the WRF model is replaced with the local climate zoning scheme; Calculate the average building height, the standard deviation of the average building height, the average building width and the average street width for each built-up area class based on the urban morphology parameters for each grid cell in each built-up area class; The average building height, the average building height standard deviation, the average building width and the average street width of each built-up area class are input into the parameter file of the WRF model to obtain the target WRF model.

[0008] Optionally, simulating weather using the target WRF model, wherein the target WRF model outputs a simulation result, specifically includes: simulating weather using the target WRF model, the target WRF model outputting simulation results, and extracting simulated air temperature, simulated relative humidity, and simulated wind speed from the simulation results; The simulated air temperature, the simulated relative humidity and the simulated wind speed are imported into the GIS platform, and the simulated air temperature, the simulated relative humidity and the simulated wind speed are interpolated to the resolution of the local climate zone using the resampling function. Figure 1 on a consistent grid.

[0009] Optionally, the step of respectively establishing historical microclimate data sets corresponding to air temperature, relative humidity and wind speed, and using a plurality of the historical microclimate data sets to train and test an XGBoost model to obtain a target XGBoost model specifically includes: Acquire monitoring data corresponding to a historical period from a meteorological monitoring station, and calculate the monitoring temperature, monitoring relative humidity and monitoring wind speed corresponding to the historical period based on the monitoring data; The monitored air temperature, the monitored relative humidity and the monitored wind speed are used as labels of the historical microclimate data sets corresponding to the air temperature, the relative humidity and the wind speed, respectively; Obtaining historical simulation results corresponding to the historical period output by the target WRF model, wherein the historical simulation results include historical simulated air temperature, historical simulated relative humidity, and historical simulated wind speed; The historical simulated air temperature, the historical simulated relative humidity, the historical simulated wind speed, and the corresponding historical periods and urban morphological parameters are used as features of the historical microclimate data set corresponding to the air temperature, relative humidity, and wind speed, respectively; Dividing the plurality of historical microclimate data sets into training sets and test sets respectively, to obtain a plurality of training sets and a plurality of test sets; Using the plurality of training sets to train an XGBoost model to obtain an intermediate XGBoost model; The intermediate XGBoost model is tested using a plurality of the test sets, the performance of the intermediate XGBoost model is evaluated using an accuracy index, and the parameters of the intermediate XGBoost model are adjusted according to the performance of the intermediate XGBoost model until the performance of the intermediate XGBoost model meets the requirements, thereby obtaining a target XGBoost model.

[0010] Optionally, inputting the simulation result to be evaluated into the target XGBoost model, wherein the target XGBoost model outputs a target temperature, a target relative humidity, and a target wind speed, specifically includes: Obtaining simulation results to be evaluated corresponding to the period to be evaluated output by the target WRF model, wherein the simulation results to be evaluated include simulated air temperature to be evaluated, simulated relative humidity to be evaluated, and simulated wind speed to be evaluated; Input the simulated temperature to be evaluated, the simulated relative humidity to be evaluated, the simulated wind speed to be evaluated, and the corresponding period to be evaluated and urban morphology parameters into the target XGBoost model; The target XGBoost model outputs a target temperature, a target relative humidity, and a target wind speed corresponding to the period to be evaluated for which the accuracy is higher than a set threshold; Among them, multiple periods to be evaluated constitute a stage to be evaluated.

[0011] Optionally, the calculating of the thermal comfort evaluation index of the built-up area according to the target air temperature, the target relative humidity and the target wind speed specifically includes: Calculate the average target temperature, average target relative humidity and average target wind speed corresponding to the stage to be evaluated according to the target air temperature, target relative humidity and target wind speed corresponding to the multiple periods to be evaluated in the stage to be evaluated output by the target XGBoost model; According to the average target air temperature, the average target relative humidity and the average target wind speed corresponding to the stage to be evaluated of each grid unit in the built-up area, the thermal comfort evaluation index NET corresponding to the stage to be evaluated of each grid unit in the built-up area is calculated: ; in, is the average target temperature, is the average target relative humidity, is the average target wind speed.

[0012] To achieve the above-mentioned purpose of the invention, the present invention also provides a refined evaluation system for thermal comfort in built-up areas based on WRF and XGBoost coupling simulation, and the refined evaluation system for thermal comfort in built-up areas based on WRF and XGBoost coupling simulation includes: Urban morphology parameter calculation module: used to calculate the urban morphology parameters of each grid unit in the built-up area based on the local climate zoning map using geospatial information data; A target WRF model generation module is used to improve the WRF model by using the local climate zoning map and a plurality of the urban morphology parameters to obtain a target WRF model; A simulation result output module: used to simulate weather using the target WRF model, wherein the target WRF model outputs simulation results, wherein the simulation results include historical simulation results and simulation results to be evaluated, and both the historical simulation results and the simulation results to be evaluated include simulated temperature, simulated relative humidity and simulated wind speed; Target XGBoost model generation module: used to establish historical microclimate data sets corresponding to air temperature, relative humidity and wind speed respectively, and train and test the XGBoost model using multiple historical microclimate data sets to obtain the target XGBoost model; Target result output module: used to input the simulation result to be evaluated into the target XGBoost model, and the target XGBoost model outputs the target temperature, target relative humidity and target wind speed; Thermal comfort evaluation index calculation module: used to calculate the thermal comfort evaluation index of the built-up area according to the target air temperature, the target relative humidity and the target wind speed.

[0013] To achieve the above-mentioned purpose of the invention, the present invention also provides a terminal, which includes: a memory, a processor, and a refined evaluation program for thermal comfort of built-up areas based on WRF and XGBoost coupling simulation, which is stored in the memory and can be run on the processor. When the refined evaluation program for thermal comfort of built-up areas based on WRF and XGBoost coupling simulation is executed by the processor, the steps of the refined evaluation method for thermal comfort of built-up areas based on WRF and XGBoost coupling simulation are implemented as described above.

[0014] To achieve the above-mentioned purpose of the invention, the present invention also provides a computer-readable storage medium, which stores a refined evaluation program for thermal comfort of built-up areas based on WRF and XGBoost coupling simulation. When the refined evaluation program for thermal comfort of built-up areas based on WRF and XGBoost coupling simulation is executed by a processor, the steps of the refined evaluation method for thermal comfort of built-up areas based on WRF and XGBoost coupling simulation are implemented as described above.

[0015] In the present invention, based on the local climate zoning map, the urban morphological parameters of each grid unit in the built-up area are calculated using geographic spatial information data; the WRF model is improved using the local climate zoning map and a plurality of the urban morphological parameters to obtain a target WRF model; the target WRF model is used to simulate weather, and the target WRF model outputs simulation results, wherein the simulation results include historical simulation results and simulation results to be evaluated, and the historical simulation results and the simulation results to be evaluated both include simulated air temperature, simulated relative humidity and simulated wind speed; historical microclimate data sets corresponding to air temperature, relative humidity and wind speed are established respectively, and an XGBoost model is trained and tested using a plurality of the historical microclimate data sets to obtain a target XGBoost model; the simulation results to be evaluated are input into the target XGBoost model, and the target XGBoost model outputs a target air temperature, a target relative humidity and a target wind speed; according to the target air temperature, the target relative humidity and the target wind speed, a thermal comfort evaluation index of the built-up area is calculated. Based on the impact mechanism of urban morphology on microclimate change at the block scale, the present invention constructs urban morphology parameters, couples the LCZ (Local Climate Zones) scheme and the machine learning model XGBoost with the WRF model, and introduces a comprehensive thermal comfort evaluation index involving temperature, relative humidity and wind speed. It realizes multi-factor, large-scale, high-precision, and high-temporal and spatial resolution thermal comfort evaluation of built-up areas, which is helpful to deeply explore the spatiotemporal distribution characteristics of urban thermal comfort and formulate urban planning and management measures to alleviate thermal discomfort. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a flow chart of a preferred embodiment of the method for refined evaluation of thermal comfort in built-up areas based on WRF and XGBoost coupling simulation of the present invention; Figure 2 It is another flow chart of a preferred embodiment of the method for refined evaluation of thermal comfort in built-up areas based on WRF and XGBoost coupling simulation of the present invention; Figure 3 It is a structural diagram of a preferred embodiment of the refined evaluation system for thermal comfort in built-up areas based on WRF and XGBoost coupling simulation of the present invention; Figure 4 It is a structural diagram of a preferred embodiment of the terminal of the present invention. DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solution and advantages of the present invention clearer and more specific, the present invention is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0018] Thermal comfort in built-up areas refers to the degree of comfort people experience with local climate conditions (such as temperature, humidity, wind speed, and solar radiation) in the urban built environment. Good urban thermal comfort plays an important role in optimizing urban spatial planning and layout, responding to climate change challenges, improving the quality of life of residents, and stimulating urban vitality. The evaluation of thermal comfort in built-up areas helps to explore the temporal and spatial variation patterns of thermal comfort in built-up areas, identify high-risk areas with poor thermal comfort conditions, and provide a scientific basis for formulating improvement strategies.

[0019] At present, the indicators for thermal comfort assessment of large-scale built-up areas are mainly calculated by linear equations constructed based on meteorological variables. These linear equations are mainly constructed by temperature (such as air temperature, surface temperature) or temperature and humidity (such as heat index, HI), but lack consideration of wind speed variables. Existing studies have shown that wind speed is one of the important factors affecting thermal comfort, especially in coastal cities with frequent sea breezes. When other climatic conditions are the same, suitable wind speed helps to avoid the feeling of stuffiness or dryness for outdoor people. Therefore, it is necessary to take wind speed into consideration when conducting refined thermal comfort assessment of built-up areas. However, due to the availability of data, wind speed variables are difficult to apply to thermal comfort assessment indicators. Traditional meteorological data acquisition methods are generally fixed-point or mobile meteorological monitoring methods, such as meteorological monitoring stations, portable observation instruments, and remote sensing observations. These methods are highly dependent on equipment deployment and manual maintenance, and the temporal and spatial resolution is difficult to adjust according to the needs of thermal comfort assessment. In addition, large-scale and long-term data collection often costs a lot of money.

[0020] Technologies such as numerical weather prediction models and machine learning models have significant advantages in simulating and generating large-scale meteorological data. Numerical weather prediction models are based on aerodynamics and thermodynamics, and use numerical methods to simulate the evolution of atmospheric states over time, thereby simulating and predicting weather conditions. Common numerical weather prediction models, such as WRF, can simulate a variety of meteorological variables with spatial resolutions of up to 100 meters and temporal resolutions of up to minutes. However, in urban areas with complex underlying surfaces, the simulation performance of such models is often poor, especially in wind speed simulation. This is because the model simplifies the land cover type of the city into a single category (i.e., impervious surface), which fails to fully consider the complex impact of urban morphology on microclimate. To address this problem, there are attempts to improve the simulation accuracy by combining multi-source data (such as LCZ maps) and urban canopy models with numerical weather prediction models, but the accuracy of wind speed is still difficult to meet the requirements of thermal comfort assessment. Machine learning models, such as random forests and XGBoost, extract knowledge from historical data by simulating the human learning process, find the rules or mapping relationships between data, and then simulate and predict new data. However, these models also have major limitations: 1) they rely on monitoring data as label data, and their simulation accuracy is poor in areas where monitoring sites are insufficient or far away from monitoring sites; 2) they use a large amount of static feature data (mainly geospatial data and urban morphology data) to train the models, and their ability to simulate sudden climate change events is insufficient. In general, although these models have the potential to generate meteorological data for refined thermal comfort assessment, a universal and reliable method has not yet been formed.

[0021] In order to solve the above technical problems, the present invention provides a refined evaluation method for thermal comfort of built-up areas based on WRF and XGBoost coupling simulation, based on a local climate zoning map, using geographic spatial information data to calculate the urban morphological parameters of each grid unit in the built-up area; using the local climate zoning map and multiple urban morphological parameters to improve the WRF model to obtain a target WRF model; using the target WRF model to simulate weather, the target WRF model outputs simulation results, wherein the simulation results include historical simulation results and simulation results to be evaluated, and the historical simulation results and the simulation results to be evaluated both include simulated temperature, simulated relative humidity and simulated wind speed; respectively establish historical microclimate data sets corresponding to temperature, relative humidity and wind speed, use multiple historical microclimate data sets to train and test an XGBoost model, and obtain a target XGBoost model; the simulation results to be evaluated are input into the target XGBoost model, and the target XGBoost model outputs a target temperature, a target relative humidity and a target wind speed; according to the target temperature, the target relative humidity and the target wind speed, the thermal comfort evaluation index of the built-up area is calculated. Based on the impact mechanism of urban morphology on microclimate change at the block scale, the present invention constructs urban morphology parameters, couples the LCZ scheme and the machine learning model XGBoost with the WRF model, and introduces a comprehensive thermal comfort evaluation index involving temperature, relative humidity and wind speed. It realizes multi-factor, large-scale, high-precision, and high-temporal and spatial resolution thermal comfort evaluation of built-up areas, which is helpful to deeply explore the spatiotemporal distribution characteristics of urban thermal comfort and formulate urban planning and management measures to alleviate thermal discomfort.

[0022] The application content is further explained below through the description of embodiments in conjunction with the accompanying drawings.

[0023] A preferred embodiment of the method for fine-grained evaluation of thermal comfort in built-up areas based on WRF and XGBoost coupling simulation of the present invention is as follows: Figure 1 and Figure 2 As shown, specifically including: S1: Based on the local climate zoning map, the urban morphological parameters of each grid cell in the built-up area are calculated using geospatial information data.

[0024] In an implementation of this embodiment, the urban morphological parameters of each grid unit in the built-up area are calculated based on the local climate zoning map using the geographic spatial information data, specifically including: S11: Acquire an LCZ map and geospatial information data, wherein the geospatial information data includes building data, land cover data, and terrain elevation data; S12: superimposing the building data, the land cover data and the terrain elevation data onto the LCZ map through a GIS (Geographic Information System) software platform, and calculating the urban morphological parameters of each grid cell in each built-up area class; Among them, the urban morphology parameters include building coverage, building height, building height standard deviation, building width, total building volume, total building area, street width, sky visibility, permeability, water coverage and altitude.

[0025] Specifically, LCZ map data and building data (i.e. Figure 2 The building vector data in the LCZ map), land cover data and terrain elevation data. Through the GIS software platform, all built-up area classes within the scope of the study city are extracted from the LCZ map data, and the urban morphological parameters UMPs are calculated for these grid cells (i.e., raster cells). In the GIS platform, the building data, land cover data and terrain elevation data are superimposed on the LCZ map, and the UMPs of each raster cell of the extracted built-up area class LCZ are calculated using a two-step raster method. Among them, the building data in vector format are rasterized with a resolution of 1m, and the raster value is the height of the building. To ensure consistent resolution, the land cover data and terrain elevation data are also resampled to a resolution of 1m. Using the raster cells on the LCZ map, various urban morphological parameters are statistically analyzed.

[0026] The definitions and calculation methods of various urban morphological parameters include: The building cover fraction (BSF) indicates the proportion of the ground surface covered by buildings within a grid cell: ; Where n is the number of buildings in the grid unit, i represents the i-th building, is the area of ​​the ground covered by the i-th building, is the area of ​​the grid cell; MBH represents the average height of buildings within a grid cell: ; in, is the height of the i-th building; The standard deviation of building height (SBH) represents the standard deviation of building height within a grid cell: ; Building width (MBW) represents the average width of a building within a grid cell: ; The gross building volume (BV) represents the sum of the heights of all buildings within a grid cell: ; The gross building area (GFA) represents the sum of the areas of all buildings within a grid cell: ; Street width (MSW) represents the average width of streets within a grid cell: ; in, is the resolution of the grid cell (100m); The sky visibility fraction (SVF) represents the shadowing effect of buildings within a grid cell: ; The water permeability fraction (PSF) indicates the area ratio of permeable water surfaces (such as green space and water bodies) within a grid cell: ; in, is the total area of ​​permeable water surface within the grid cell; The water body coverage fraction (WSF) represents the area ratio of water bodies within a grid cell: ; in, is the total area of ​​the water body within the grid cell; Altitude (MA) represents the average altitude of the terrain within a grid cell: ; in, Terrain Grid altitude.

[0027] S2: improving the WRF model using the local climate zoning map and a plurality of the urban morphology parameters to obtain a target WRF model.

[0028] In an implementation of this embodiment, the improving the WRF model by using the local climate zoning map and the plurality of urban morphology parameters to obtain the target WRF model specifically includes: S21: Setting the simulation period, nested domain and projection coordinates of the WRF model; S22: coupling the local climate zoning map with the land cover data of the lowest nested domain of the WRF model, and replacing the land cover classification scheme of the lowest nested domain of the WRF model with the local climate zoning scheme; S23: calculating the average building height, the average building height standard deviation, the average building width and the average street width of each built-up area class according to the urban morphological parameters of each grid cell in each built-up area class; S24: Input the average building height, the average building height standard deviation, the average building width and the average street width of each built-up area class into the parameter file of the WRF model to obtain a target WRF model.

[0029] Specifically, the WRF model (i.e. Figure 2 The workflow of the WRF model in the WRF model includes three parts: pre-processing, main program and post-processing. The pre-processing stage mainly includes setting the simulation period, nested domains and projection coordinates. In addition, in order to couple the LCZ scheme, the land cover data of the lowest nested domain (referring to the land cover data that comes with the WRF model) needs to be replaced with the LCZ map. At the same time, the average values ​​of the four UMPs of building height, building height standard deviation, building width and street width of each type of LCZ (built-up area) are calculated and input into the LCZ parameter file of the WRF model. In the main program stage, physical parameterization schemes need to be configured. These schemes include microphysical schemes, long-wave radiation schemes and short-wave radiation schemes, etc., to accurately describe the atmospheric physical processes in the study area. The post-processing stage is mainly responsible for extracting the simulation results, including the three key variables of temperature, relative humidity and wind speed (i.e. Figure 2 Hourly simulation data of temperature, humidity and wind in the

[0030] S3: simulating weather using the target WRF model, wherein the target WRF model outputs simulation results, wherein the simulation results include historical simulation results and simulation results to be evaluated, and both the historical simulation results and the simulation results to be evaluated include simulated air temperature, simulated relative humidity and simulated wind speed.

[0031] In an implementation of this embodiment, the target WRF model is used to simulate weather, and the target WRF model outputs a simulation result, which specifically includes: S31: simulating weather using the target WRF model, the target WRF model outputting simulation results, and extracting simulated temperature, simulated relative humidity and simulated wind speed from the simulation results; S32: Import the simulated air temperature, the simulated relative humidity and the simulated wind speed into the GIS platform, and use the resampling function to interpolate the simulated air temperature, the simulated relative humidity and the simulated wind speed to a resolution consistent with the local climate zone. Figure 1 on a consistent grid.

[0032] Specifically, the present invention uses the improved target WRF model for numerical weather forecasting, and the historical simulation results corresponding to the historical simulation period (i.e., the historical period) output by the target WRF model are used in the training XGBoost model stage. In the built-up area thermal comfort assessment stage, the target XGBoost model is input with the simulation results to be assessed corresponding to the simulation period to be assessed (i.e., the period to be assessed) output by the target WRF model. It should be noted that the period to be assessed can be a historical period or a future period. In other words, the period to be assessed can be the same period as the historical period, that is, the historical simulation results used to train the XGBoost model can also be used as the input of the target XGBoost model in the thermal comfort assessment of the built-up area in the historical period.

[0033] S4: Establish historical microclimate data sets corresponding to air temperature, relative humidity and wind speed respectively, and use multiple historical microclimate data sets to train and test the XGBoost model to obtain a target XGBoost model.

[0034] In one implementation of this embodiment, the establishment of historical microclimate data sets corresponding to air temperature, relative humidity and wind speed, respectively, and the use of multiple historical microclimate data sets to train and test the XGBoost model to obtain the target XGBoost model specifically include: S41: Acquire monitoring data corresponding to a historical period from a meteorological monitoring station, and calculate a monitoring temperature, a monitoring relative humidity, and a monitoring wind speed corresponding to the historical period based on the monitoring data; S42: using the monitored air temperature, the monitored relative humidity, and the monitored wind speed as labels of historical microclimate data sets corresponding to air temperature, relative humidity, and wind speed, respectively; S43: Obtaining historical simulation results corresponding to the historical period output by the target WRF model, wherein the historical simulation results include historical simulated air temperature, historical simulated relative humidity and historical simulated wind speed; S44: using the historical simulated air temperature, the historical simulated relative humidity, the historical simulated wind speed, and the corresponding historical periods and urban morphological parameters as features of the historical microclimate data set corresponding to air temperature, relative humidity, and wind speed, respectively; S45: Dividing the plurality of historical microclimate data sets into training sets and test sets respectively, to obtain a plurality of training sets and a plurality of test sets; S46: Using the plurality of training sets to train an XGBoost model to obtain an intermediate XGBoost model; S47: Use multiple test sets to test the intermediate XGBoost model, use accuracy indicators to evaluate the performance of the intermediate XGBoost model, and adjust the parameters of the intermediate XGBoost model according to the performance of the intermediate XGBoost model until the performance of the intermediate XGBoost model meets the requirements, thereby obtaining a target XGBoost model.

[0035] Specifically, firstly, historical microclimate datasets were established for the three variables of temperature, relative humidity and wind speed (i.e. Figure 2 The machine learning datasets in the XGBoost model are used to train the XGBoost model. Each historical microclimate dataset is based on the measured data corresponding to the historical period of the meteorological monitoring station (i.e. Figure 2 The meteorological monitoring data in the dataset is used as labels. In terms of feature selection, three types of data are included: WRF simulation results (referring to historical simulation results), urban morphology parameters UMPs, and simulation time (i.e. simulation period, referring to the historical period). Subsequently, the importance of each feature is evaluated, and the key features that contribute more to the total importance are selected for model training. The historical microclimate dataset is divided into a training set and a test set according to a certain ratio, where the training set is used to train the XGBoost model and the test set is used to test the performance of the XGBoost model. The accuracy index is used to evaluate the accuracy of the XGBoost model for the test set, so as to adjust parameters such as n_estimators, learning_rate, and max_depth to improve the learning ability of the XGBoost model.

[0036] S5: Inputting the simulation results to be evaluated into the target XGBoost model, and the target XGBoost model outputs a target air temperature, a target relative humidity, and a target wind speed.

[0037] In one implementation of this embodiment, the inputting of the simulation result to be evaluated into the target XGBoost model, wherein the target XGBoost model outputs a target temperature, a target relative humidity, and a target wind speed, specifically includes: S51: Obtaining simulation results to be evaluated corresponding to the period to be evaluated output by the target WRF model, wherein the simulation results to be evaluated include simulated air temperature to be evaluated, simulated relative humidity to be evaluated, and simulated wind speed to be evaluated; S52: inputting the simulated temperature to be evaluated, the simulated relative humidity to be evaluated, the simulated wind speed to be evaluated, and the corresponding period to be evaluated and urban morphological parameters into the target XGBoost model; S53: the target XGBoost model outputs a target temperature, a target relative humidity, and a target wind speed corresponding to the period to be evaluated whose accuracy is higher than a set threshold; Among them, multiple periods to be evaluated constitute a stage to be evaluated.

[0038] Specifically, first determine the simulation stage to be evaluated (i.e., the stage to be evaluated, referring to the month to be evaluated) and the multiple simulation periods to be evaluated in the simulation stage to be evaluated (i.e., the period to be evaluated, referring to the hour to be evaluated), use the target WRF model to obtain the simulation results to be evaluated corresponding to the multiple periods to be evaluated, and input the simulation results to be evaluated, the period to be evaluated and the urban morphological parameters into the trained target XGBoost model, and the target XGBoost model generates a high-precision target microclimate data set, that is, outputs the target temperature, target relative humidity and target wind speed corresponding to the multiple periods to be evaluated. It should be noted that the unit of the simulation stage is month, and the unit of the simulation period is hour. The present invention obtains a microclimate data set with multiple factors, large range, high precision and high spatiotemporal resolution by combining the technical advantages of the numerical weather forecast model WRF and the machine learning model XGBoost, thereby supporting the refined evaluation of thermal comfort in the built-up area, wherein the time accuracy of the microclimate data set is 1 hour and the spatial accuracy is 100m.

[0039] It should be noted that the simulation results output by the target WRF model are the initial simulation results, while the simulation results output by the target XGBoost model are the final simulation results, namely the target simulation results. The target simulation results are more accurate than the initial simulation results, which helps to achieve a refined assessment of thermal comfort in built-up areas.

[0040] S6: Calculate a thermal comfort assessment index for the built-up area based on the target air temperature, the target relative humidity and the target wind speed.

[0041] In an implementation of this embodiment, the calculation of the thermal comfort evaluation index of the built-up area according to the target air temperature, the target relative humidity and the target wind speed specifically includes: S61: Calculate the average target temperature, average target relative humidity and average target wind speed corresponding to the stage to be evaluated according to the target air temperature, the target relative humidity and the target wind speed corresponding to the multiple periods to be evaluated in the stage to be evaluated output by the target XGBoost model; S62: Calculate the thermal comfort evaluation index NET corresponding to the to-be-evaluated stage for each grid unit in the built-up area according to the average target air temperature, the average target relative humidity and the average target wind speed corresponding to the to-be-evaluated stage for each grid unit in the built-up area: ; in, is the average target temperature, is the average target relative humidity, is the average target wind speed.

[0042] Specifically, for the high-precision target microclimate data set generated by the target XGBoost model, the average target temperature, average target relative humidity and average target wind speed of the stage to be evaluated are calculated respectively, and the overall accuracy of each meteorological element is evaluated. When the NET value is between 17 and 21, the built-up area is in a comfortable state; when the NET value is greater than 21, the built-up area is in a hot uncomfortable state (when the NET value is between 21 and 23, the built-up area is in a slightly hot state; when the NET value is between 23 and 27, the built-up area is in a hot state; when the NET value is above 27, the built-up area is in a hot state); when the NET value is lower than 17, the built-up area is in a cold uncomfortable state (when the NET value is between 9 and 17, the built-up area is in a slightly cold state; when the NET value is between 1 and 9, the built-up area is in a cold state). The present invention optimizes the WRF simulation results from two aspects: coupling the LCZ scheme and constructing a machine learning model, and finally carries out the thermal comfort evaluation of the built-up area through the indicator NET.

[0043] Taking City A as an example, thermal comfort assessment of built-up areas was carried out for four months in the dry and wet seasons. Figure 2 As shown, the specific steps include: Step 1: Calculation of urban morphology parameters.

[0044] A standard grid was established for City A, and UMPs were calculated for grid cells (i.e., grid cells) within the built-up area. Based on previous studies, 11 UMPs were selected: building coverage, building height, standard deviation of building height, building width, total building volume, total building area, street width, sky visibility, permeability, water coverage, and altitude. LCZ map data, building data, land cover data, and terrain elevation data were obtained from a third-party public open platform. All built-up LCZs within the scope of City A were extracted from the LCZ map data using the GIS software platform, and UMPs were calculated for these grid cells. In the GIS platform, data such as buildings were superimposed on the LCZ map, and the UMPs were calculated for the extracted built-up LCZs using a two-step grid method. Building data in vector format were rasterized with a resolution of 1 m, and the grid value was the height of the building. To ensure consistent resolution, land cover data and terrain elevation data were also resampled to a resolution of 1 m. Using these grid cells, various urban morphological parameters were statistically analyzed.

[0045] Step 2: WRF simulation of coupled LCZ scheme.

[0046] According to the climate characteristics of City A, two months were selected from the dry season and the wet season, and one week was selected in each month as both the historical simulation period and the simulation period to be evaluated (that is, in this embodiment, the historical period and the period to be evaluated are the same period), and a total of 4 WRF simulations were carried out. In order to obtain better simulation results, the selected periods are mainly calm and stable weather. Since the WRF model has a warm-up period after starting to run, the specific embodiment only uses the simulation results from the second to the seventh day of each week, as shown in Table 1: Table 1: Simulation schedule

[0047] Four nested domains were constructed around the study area of ​​City A. The grid spacing of the nested domains from the outermost to the innermost layer was 9, 3, 1, and 0.333 km, respectively. The outermost nested domain covers an area of ​​1170×954 km. The innermost nested domain is centered on the study area and covers an area of ​​121×75 km. The coordinate system of the WRF model uses the Mercator projection. The LCZ map is coupled with the land cover data of the nested domain using the Python-based w2w module. After coupling, the land cover classification scheme (MODIS classification) of the innermost nested domain is replaced by the LCZ scheme. For each type of built-up area LCZ, the average MBH, SBH, MBW, and MSW are calculated, and the results are input into the WRF model.

[0048] The initial and boundary conditions of the WRF model are derived from the global data assimilation system. The horizontal resolution of these data is 0.25°×0.25° and they are updated every 6 hours. In terms of physical parameter schemes, the Lin microphysics scheme, the RRTM (Rapid Radiative Transfer Model) longwave radiation scheme, the Dudhia shortwave radiation scheme, the YSU planetary boundary layer scheme, and the Noah-MP land surface scheme are selected.

[0049] After completing the simulation of the four time periods, the hourly simulated data of 2m temperature, 2m relative humidity, and 10m wind speed were extracted from the result file of the WRF model. The simulated data were imported into the GIS platform and interpolated to a 100m resolution grid using the resampling function (the resolution is the same as that of the LCZ map). Figure 1 to).

[0050] Step 3: XGBoost model construction.

[0051] For the three climate variables of temperature, relative humidity and wind speed, hourly meteorological simulation datasets, namely historical microclimate datasets (including samples), are constructed for training XGBoost models. In terms of the labels of the historical microclimate datasets, the monitoring data corresponding to the historical period of the meteorological monitoring station in City A are used as references. These monitoring data are generated by several meteorological monitoring stations located in the built-up area, and are published on the official website of the Meteorological Bureau of City A and updated every minute. Using these monitoring data, the hourly monitored temperature, monitored relative humidity and monitored wind speed are calculated as labels. In terms of the characteristics of the historical microclimate dataset, there are three types of data, including WRF simulation results (historical simulation results), UMPs and simulation periods (historical periods). Among them, the WRF historical simulation results are the resampled simulated temperature, simulated relative humidity and simulated wind speed. All 11 UMPs are used as dataset features. In addition, the simulation time is reflected by hours and months. Since hours and months are periodic, hours and months are respectively and month The sine and cosine values ​​are converted to generate 4 time features in total. The conversion formula is: ; ; ; ; in, , , and All are time characteristics. (0, 1, 2, ..., 23) is the number of hours of the sample on the simulated day, is the simulated month where the sample is located. The three historical microclimate datasets finally obtained each have 66,816 samples and 18 features.

[0052] After determining the historical microclimate dataset, key features and parameter tuning were further selected. The permutation importance provided by the Python-based scikit-learn extension package was used to evaluate the importance of features. The XGBoost model was trained 10 times with the average of the permutation importance as the importance of the feature. The features were sorted from high to low by permutation importance, and key features that contributed more than 98% to the total importance were selected for training the model. For each historical microclimate dataset, 70% of the samples were selected as the training set and the remaining 30% as the test set. The root mean square error (RMSE), mean absolute error (MAE), and mean percentage error (MAPE) were used to evaluate the training accuracy of the XGBoost model. According to the performance of the XGBoost model, the matrix search method was used to adjust parameters such as n_estimators, learning_rate, and max_depth.

[0053] Using the trained XGBoost model, the simulated data (temperature, relative humidity, wind speed) output by the WRF model, the simulation period and urban morphological parameters are used as input to generate the target microclimate datasets of temperature, relative humidity and wind speed. The target microclimate dataset covers a total of 24 days, with a time accuracy of 1 hour and a spatial accuracy of 100m, covering all built-up areas in City A.

[0054] Step 4: Thermal comfort assessment.

[0055] Using the generated target microclimate dataset, the microclimate assessment, i.e. thermal comfort assessment, of the built-up area of ​​City A was carried out in different months of the dry and wet seasons (i.e. the simulation period selected in step 1). The average values ​​of hourly temperature, hourly relative humidity, and hourly wind speed were calculated for each month, and the root mean square error (RMSE), mean absolute error (MAE), and mean percentage error (MAPE) were used to evaluate the accuracy of these assessment data. Based on the average values ​​of hourly temperature, hourly relative humidity, and hourly wind speed for each month, the average NET values ​​of the built-up area of ​​City A in February, May, July, and November were calculated, and their corresponding thermal comfort states were evaluated.

[0056] The present invention proposes a refined assessment method for thermal comfort in built-up areas based on WRF and XGBoost coupling simulation. The advantages of the present invention are that it has many climate factors, a large assessment range, high temporal and spatial resolution, and high data accuracy. Specifically: 1) The thermal comfort evaluation index NET incorporates air temperature, relative humidity and wind speed. Compared with the common index that only considers temperature and humidity, the index of the present invention can more comprehensively reflect the human body's perception of the complex outdoor thermal environment, and is especially suitable for areas with a greater impact of the wind environment; 2) Compared with the traditional evaluation method which is limited by the scope and number of meteorological monitoring stations, the present invention can more easily realize large-scale thermal comfort evaluation at the city and even regional scales, which is conducive to the efficient implementation of large-scale thermal comfort research; 3) Based on the simulation performance of the WRF model, the present invention can achieve a time resolution of minutes and a spatial resolution of hundreds of meters, which can reach the temporal and spatial resolution level of meteorological monitoring stations, and is suitable for microclimate research at the block scale; 4) The present invention uses urban morphological parameters to couple the numerical simulation of atmospheric diffusion with machine learning methods, which can achieve high-precision microclimate simulation, especially for the wind speed variable, which is the most difficult to simulate. The evaluation accuracy can reach nearly 90%, which is significantly better than the existing methods (usually below 70%).

[0057] In addition, based on the above-mentioned refined evaluation method of thermal comfort in built-up areas based on WRF and XGBoost coupling simulation, the present invention also provides a refined evaluation system of thermal comfort in built-up areas based on WRF and XGBoost coupling simulation, wherein a preferred embodiment of the refined evaluation system of thermal comfort in built-up areas based on WRF and XGBoost coupling simulation is as follows: Figure 3 As shown, specifically including: Urban morphology parameter calculation module 01: used to calculate the urban morphology parameters of each grid unit in the built-up area based on the local climate zoning map using geographic spatial information data; Target WRF model generation module 02: used to improve the WRF model by using the local climate zoning map and the urban morphology parameters to obtain a target WRF model; Simulation result output module 03: used to simulate weather using the target WRF model, the target WRF model outputs simulation results, wherein the simulation results include historical simulation results and simulation results to be evaluated, and the historical simulation results and the simulation results to be evaluated both include simulated temperature, simulated relative humidity and simulated wind speed; Target XGBoost model generation module 04: used to respectively establish historical microclimate data sets corresponding to air temperature, relative humidity and wind speed, and train and test the XGBoost model using a plurality of the historical microclimate data sets to obtain a target XGBoost model; Target result output module 05: used to input the simulation result to be evaluated into the target XGBoost model, and the target XGBoost model outputs the target temperature, target relative humidity and target wind speed; Thermal comfort evaluation index calculation module 06: used to calculate the thermal comfort evaluation index NET of the built-up area according to the target air temperature, the target relative humidity and the target wind speed.

[0058] In addition, based on the above-mentioned method and system for refined evaluation of thermal comfort in built-up areas based on WRF and XGBoost coupling simulation, the present invention also provides a terminal accordingly, wherein a preferred embodiment of the terminal is as follows: Figure 4 As shown, it specifically includes a processor 10, a memory 20 and a display 30. Figure 4 Only some components of the terminal are shown, but 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.

[0059] In some embodiments, the memory 20 may be an internal storage unit of the terminal, such as a hard disk or memory of the terminal. In other embodiments, the memory 20 may also be an external storage device of the terminal, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, and a flash card (Flash Card) equipped on the terminal. Further, the memory 20 may also include both an internal storage unit of the terminal and an external storage device. The memory 20 is used to store application software and various types of data installed on the terminal, such as program codes of the storage terminal. The memory 20 may also be used to temporarily store data that has been output or is to be output. In one embodiment, the memory 20 stores a refined evaluation program 40 for thermal comfort in built-up areas based on WRF and XGBoost coupling simulation, and the refined evaluation program 40 for thermal comfort in built-up areas based on WRF and XGBoost coupling simulation can be executed by the processor 10, thereby realizing the steps of the refined evaluation method for thermal comfort in built-up areas based on WRF and XGBoost coupling simulation in the present application.

[0060] In some embodiments, the processor 10 may be a central processing unit (CPU), a microprocessor or other data processing chip, used to run program codes or process data stored in the memory 20, such as executing a refined evaluation program 40 for thermal comfort in built-up areas based on WRF and XGBoost coupling simulation.

[0061] In some embodiments, the display 30 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch device, etc. The display 30 is used to display information on the terminal and to display a visual user interface.

[0062] In one embodiment, when the processor 10 executes the refined evaluation program 40 of built-up area thermal comfort based on WRF and XGBoost coupling simulation in the memory 20, the steps of the refined evaluation method of built-up area thermal comfort based on WRF and XGBoost coupling simulation as described above are implemented.

[0063] The present invention also provides a computer-readable storage medium accordingly, wherein the computer-readable storage medium stores a refined evaluation program for thermal comfort of built-up areas based on WRF and XGBoost coupling simulation, and when the refined evaluation program for thermal comfort of built-up areas based on WRF and XGBoost coupling simulation is executed by a processor, the steps of the refined evaluation method for thermal comfort of built-up areas based on WRF and XGBoost coupling simulation are implemented as described above.

[0064] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the existence of other identical elements in the process, method, article or terminal including the element.

[0065] Of course, those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing related hardware (such as a processor, a controller, etc.) through a computer program, and the program can be stored in a computer-readable storage medium that can be read by a computer, and the program can include the processes of the above-mentioned method embodiments when executed. The computer-readable storage medium can be a memory, a disk, an optical disk, etc.

[0066] It should be understood that the application of the present invention is not limited to the above examples. For ordinary technicians in this field, improvements or changes can be made based on the above description. All these improvements and changes should fall within the scope of protection of the claims attached to the present invention.

Claims

1. A refined evaluation method for thermal comfort in built-up areas based on WRF and XGBoost coupling simulation, characterized in that: The refined evaluation method of thermal comfort in built-up areas based on WRF and XGBoost coupling simulation includes: Based on the local climate zoning map, the urban morphological parameters of each grid unit in the built-up area are calculated using geospatial information data; Improve the WRF model using the local climate zoning map and a plurality of the urban morphology parameters to obtain a target WRF model; Simulating weather using the target WRF model, wherein the target WRF model outputs simulation results, wherein the simulation results include historical simulation results and simulation results to be evaluated, and both the historical simulation results and the simulation results to be evaluated include simulated temperature, simulated relative humidity, and simulated wind speed; Establishing historical microclimate data sets corresponding to air temperature, relative humidity, and wind speed respectively, and using a plurality of the historical microclimate data sets to train and test an XGBoost model to obtain a target XGBoost model; Inputting the simulation result to be evaluated into the target XGBoost model, the target XGBoost model outputs a target air temperature, a target relative humidity, and a target wind speed; The thermal comfort assessment index of the built-up area is calculated based on the target air temperature, the target relative humidity and the target wind speed.

2. According to claim 1, the refined evaluation method of thermal comfort in built-up areas based on WRF and XGBoost coupling simulation is characterized in that: The method of calculating the urban morphological parameters of each grid unit in the built-up area based on the local climate zoning map using geographic spatial information data specifically includes: Obtaining a LCZ map and geospatial information data, wherein the geospatial information data includes building data, land cover data, and terrain elevation data; Through a GIS software platform, the building data, the land cover data and the terrain elevation data are superimposed on the LCZ map to calculate the urban morphological parameters of each grid cell in each built-up area class; Among them, the urban morphology parameters include building coverage, building height, building height standard deviation, building width, total building volume, total building area, street width, sky visibility, permeability, water coverage and altitude.

3. The refined evaluation method for thermal comfort in built-up areas based on WRF and XGBoost coupling simulation according to claim 2 is characterized in that: The improving the WRF model by using the local climate zoning map and the plurality of urban morphological parameters to obtain the target WRF model specifically includes: Set the simulation period, nested domain and projection coordinates of the WRF model; The local climate zoning map is coupled with the land cover data of the lowest nested domain of the WRF model, and the land cover classification scheme of the lowest nested domain of the WRF model is replaced with the local climate zoning scheme; Calculate the average building height, the standard deviation of the average building height, the average building width and the average street width for each built-up area class based on the urban morphology parameters for each grid cell in each built-up area class; The average building height, the average building height standard deviation, the average building width and the average street width of each built-up area class are input into the parameter file of the WRF model to obtain the target WRF model.

4. The refined evaluation method for thermal comfort in built-up areas based on WRF and XGBoost coupling simulation according to claim 1 is characterized in that: The target WRF model is used to simulate weather, and the target WRF model outputs simulation results, specifically including: simulating weather using the target WRF model, the target WRF model outputting simulation results, and extracting simulated air temperature, simulated relative humidity, and simulated wind speed from the simulation results; The simulated air temperature, the simulated relative humidity and the simulated wind speed are imported into the GIS platform, and the simulated air temperature, the simulated relative humidity and the simulated wind speed are interpolated to a grid with a resolution consistent with the local climate zoning map using a resampling function.

5. The refined evaluation method for thermal comfort in built-up areas based on WRF and XGBoost coupling simulation according to claim 4 is characterized in that: The method of establishing historical microclimate data sets corresponding to air temperature, relative humidity and wind speed respectively, and using a plurality of the historical microclimate data sets to train and test the XGBoost model to obtain the target XGBoost model specifically includes: Acquire monitoring data corresponding to a historical period from a meteorological monitoring station, and calculate the monitoring temperature, monitoring relative humidity and monitoring wind speed corresponding to the historical period based on the monitoring data; The monitored air temperature, the monitored relative humidity and the monitored wind speed are used as labels of the historical microclimate data sets corresponding to the air temperature, the relative humidity and the wind speed, respectively; Obtaining historical simulation results corresponding to the historical period output by the target WRF model, wherein the historical simulation results include historical simulated air temperature, historical simulated relative humidity, and historical simulated wind speed; The historical simulated air temperature, the historical simulated relative humidity, the historical simulated wind speed, and the corresponding historical periods and urban morphological parameters are used as features of the historical microclimate data set corresponding to the air temperature, relative humidity, and wind speed, respectively; Dividing the plurality of historical microclimate data sets into training sets and test sets respectively, to obtain a plurality of training sets and a plurality of test sets; Using the plurality of training sets to train an XGBoost model to obtain an intermediate XGBoost model; The intermediate XGBoost model is tested using a plurality of the test sets, the performance of the intermediate XGBoost model is evaluated using an accuracy index, and the parameters of the intermediate XGBoost model are adjusted according to the performance of the intermediate XGBoost model until the performance of the intermediate XGBoost model meets the requirements, thereby obtaining a target XGBoost model.

6. The refined evaluation method for thermal comfort in built-up areas based on WRF and XGBoost coupling simulation according to claim 4 or 5, characterized in that: The inputting of the simulation result to be evaluated into the target XGBoost model, wherein the target XGBoost model outputs a target temperature, a target relative humidity, and a target wind speed, specifically includes: Obtaining simulation results to be evaluated corresponding to the period to be evaluated output by the target WRF model, wherein the simulation results to be evaluated include simulated air temperature to be evaluated, simulated relative humidity to be evaluated, and simulated wind speed to be evaluated; Input the simulated temperature to be evaluated, the simulated relative humidity to be evaluated, the simulated wind speed to be evaluated, and the corresponding period to be evaluated and urban morphology parameters into the target XGBoost model; The target XGBoost model outputs a target temperature, a target relative humidity, and a target wind speed corresponding to the period to be evaluated for which the accuracy is higher than a set threshold; Among them, multiple periods to be evaluated constitute a stage to be evaluated.

7. The refined evaluation method for thermal comfort in built-up areas based on WRF and XGBoost coupling simulation according to claim 6 is characterized in that: The calculating of the thermal comfort evaluation index of the built-up area according to the target air temperature, the target relative humidity and the target wind speed specifically includes: Calculate the average target temperature, average target relative humidity and average target wind speed corresponding to the stage to be evaluated according to the target air temperature, target relative humidity and target wind speed corresponding to the multiple periods to be evaluated in the stage to be evaluated output by the target XGBoost model; According to the average target air temperature, the average target relative humidity and the average target wind speed corresponding to the stage to be evaluated of each grid unit in the built-up area, the thermal comfort evaluation index NET corresponding to the stage to be evaluated of each grid unit in the built-up area is calculated: ； in, is the average target temperature, is the average target relative humidity, is the average target wind speed.

8. A refined evaluation system for thermal comfort in built-up areas based on WRF and XGBoost coupling simulation, characterized in that: The refined evaluation system for thermal comfort in built-up areas based on WRF and XGBoost coupling simulation includes: Urban morphology parameter calculation module: used to calculate the urban morphology parameters of each grid unit in the built-up area based on the local climate zoning map using geospatial information data; A target WRF model generation module is used to improve the WRF model by using the local climate zoning map and a plurality of the urban morphology parameters to obtain a target WRF model; A simulation result output module: used to simulate weather using the target WRF model, wherein the target WRF model outputs simulation results, wherein the simulation results include historical simulation results and simulation results to be evaluated, and both the historical simulation results and the simulation results to be evaluated include simulated temperature, simulated relative humidity and simulated wind speed; Target XGBoost model generation module: used to establish historical microclimate data sets corresponding to air temperature, relative humidity and wind speed respectively, and train and test the XGBoost model using multiple historical microclimate data sets to obtain the target XGBoost model; Target result output module: used to input the simulation result to be evaluated into the target XGBoost model, and the target XGBoost model outputs the target temperature, target relative humidity and target wind speed; Thermal comfort evaluation index calculation module: used to calculate the thermal comfort evaluation index of the built-up area according to the target air temperature, the target relative humidity and the target wind speed.

9. A terminal, characterized in that: The terminal includes: a memory, a processor, and a refined evaluation program for thermal comfort of built-up areas based on WRF and XGBoost coupling simulation, which is stored in the memory and can be run on the processor. When the refined evaluation program for thermal comfort of built-up areas based on WRF and XGBoost coupling simulation is executed by the processor, the steps of the refined evaluation method for thermal comfort of built-up areas based on WRF and XGBoost coupling simulation are implemented as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a refined evaluation program for thermal comfort of built-up areas based on WRF and XGBoost coupling simulation. When the refined evaluation program for thermal comfort of built-up areas based on WRF and XGBoost coupling simulation is executed by a processor, the steps of the refined evaluation method for thermal comfort of built-up areas based on WRF and XGBoost coupling simulation as described in any one of claims 1 to 7 are implemented.

Images