PM2.5 concentration prediction method and device based on block morphological attenuation theory

Abstract

The invention discloses a PM2.5 concentration prediction method and device based on a block morphological attenuation theory, and is used for solving the technical problems that a traditional PM2.5 prediction model depends on supercomputation and complex parameters, does not quantify a block attenuation mechanism, lacks LCZ attribute association and cannot completely describe a diffusion rule, and finally, the prediction precision is limited. The method is suitable for high-density cities in humid and hot climate areas. The method comprises the following steps: acquiring urban form information and meteorological data of a target area, and outputting an LCZ classification map and morphological attenuation attributes of each LCZ block according to a preset LCZ type; dividing windward-side and leeward-side blocks based on meteorological data, and calculating the distance between a PM2.5 concentration moving test point of each block and a road moving source; determining a pollutant diffusion path according to the meteorological data, selecting a block in the path as a target block, and carrying out attenuation region division to obtain a plurality of levels of attenuation regions; and an initial attenuation prediction model is constructed, a target model is obtained through multi-stage iteration fitting, and PM2.5 concentration prediction values of the windward side and the leeward side are predicted and output.

Description

Technical Field

[0001] This invention relates to the field of urban environmental monitoring technology, and in particular to a PM2.5 monitoring method based on the theory of street morphological attenuation. 2.5 Concentration prediction methods and devices. Background Technology

[0002] With the rapid advancement of urbanization, urban spatial patterns are continuously expanding and being restructured, leading to an increase in fine particulate matter (PM2.5). 2.5 (Particulate Matter 2.5) Pollution has gradually evolved into a core environmental problem threatening public health and safety and hindering sustainable environmental development. From the perspective of pollution sources, mobile sources from road traffic (such as motor vehicle exhaust emissions), due to their high emission intensity and wide coverage, have become a major contributor to urban PM2.5 pollution. 2.5 One of the main contributors to excessive PM2.5 concentrations; and from the perspective of the pollution diffusion process, PM2.5 emitted by this type of pollution source... 2.5 It is not simply dispersed through the atmosphere; its transport path and attenuation rate are significantly regulated by urban spatial morphology (such as street layout and building distribution) and local climatic conditions (such as wind direction and speed), forming a complex pollution system with interconnected "emissions-diffusion-impact." This macro-environmental situation highlights the importance of PM2.5 research and development. 2.5 The necessity of accurate concentration prediction also sets the core scenario for the research and application of existing prediction methods, which need to take into account the comprehensive impact of pollution source characteristics and complex environmental factors at the urban scale.

[0003] In PM 2.5 In the field of concentration prediction, traditional prediction models have formed a technical system represented by CFD simulation (Computational Fluid Dynamics) and atmospheric chemical transport models. These models rely on refined physicochemical mechanisms to achieve high-precision calculation of pollutant diffusion processes and show advantages in concentration simulation of small-scale local areas (such as a single block).

[0004] But when facing PM at the city scale 2.5 For concentration prediction, traditional models are inefficient due to their over-reliance on supercomputing resources and complex parameterization processes, making them unsuitable for predicting PM2.5 concentrations at the urban scale. 2.5 The practical need for real-time dynamic prediction of PM concentrations has not yet been addressed by quantifying the physical mechanisms by which urban street spatial structure contributes to pollutant attenuation. In particular, there is a lack of systematic correlation analysis between the inherent attributes (such as building density and building height) and additional attributes (such as vegetation cover and impervious surface area) of local climate zones (LCZs). This prevents models from fully characterizing PM under complex urban morphologies. 2.5 The diffusion and attenuation patterns ultimately limit the accuracy of predictions. Summary of the Invention

[0005] This invention provides a PM based on the theory of street morphological attenuation. 2.5 Concentration prediction methods and devices for solving traditional PM2.5 concentration problems. 2.5 The prediction model relies on supercomputing and complex parameters, and lacks quantification of the street decay mechanism and LCZ attribute correlation, thus failing to fully characterize the diffusion law, ultimately leading to technical problems that limit prediction accuracy.

[0006] The first aspect of this invention provides a PM based on the theory of street morphological attenuation. 2.5 Concentration prediction methods include:

[0007] Acquire urban morphology information and meteorological data of the target area, and output an LCZ classification map of the target area and the morphological attenuation attributes of each LCZ block in the LCZ classification map based on multiple preset LCZ types and the urban morphology information.

[0008] Based on the meteorological data, multiple LCZ blocks in the LCZ classification map are classified, and multiple windward and leeward blocks are output.

[0009] Based on the meteorological data, the PM2.5 concentration in each of the windward and leeward sides is calculated. 2.5 The distance between the concentration movement test point and the road-based mobile source;

[0010] Based on the meteorological data, the pollutant diffusion path is determined, and the windward and leeward blocks within the pollutant diffusion path are taken as target blocks. The multiple target blocks are then divided into attenuation areas to obtain multiple windward and leeward level attenuation areas.

[0011] An initial attenuation prediction model is constructed for both the windward and leeward sides, and the morphological attenuation attributes and PM of target blocks in each of the windward and leeward attenuation level regions are used. 2.5 The distance between the concentration movement test point and the road movement source is used to perform multi-level attenuation region iterative fitting on the initial attenuation prediction model on the windward and leeward sides to determine the target attenuation prediction model on the windward and leeward sides.

[0012] The target area is analyzed for PM levels on the windward and leeward sides using the aforementioned target attenuation prediction model. 2.5 Concentration prediction, outputting the PM2.5 concentration on the windward side of the target area. 2.5 Predicted concentration values ​​and leeward PM 2.5 Predicted concentration values.

[0013] Optionally, the urban morphology information includes road network data, remote sensing images, and surface parameter data; the step of outputting an LCZ classification map of the target area and the morphological attenuation attributes of each LCZ block in the LCZ classification map based on multiple preset LCZ types and the urban morphology information includes:

[0014] The road network data is preprocessed to output the urban road network of the target area;

[0015] The urban road network of the target area is used as the LCZ block segmentation boundary. Based on the LCZ block segmentation boundary, the target area is segmented according to the remote sensing data using GIS spatial analysis methods, and multiple initial LCZ blocks are output.

[0016] Extract the morphological attenuation attributes of each initial LCZ block, and classify each initial LCZ block according to the morphological attenuation attributes of each initial LCZ block based on multiple preset LCZ types, and output multiple LCZ blocks with preset LCZ type labels.

[0017] Each LCZ block with a preset LCZ type label is verified to generate multiple LCZ blocks, and multiple LCZ blocks are used to construct an LCZ classification map of the target area.

[0018] Optionally, the windward and leeward side blocks include windward side blocks and leeward side blocks; the classification of multiple LCZ blocks in the LCZ classification map based on the meteorological data, outputting multiple windward and leeward side blocks, includes:

[0019] The meteorological data is used to divide the target area into wind-side partitions, and the windward and leeward global partition ranges of the target area are output.

[0020] Based on the global partition range on the windward side and the global partition range on the leeward side, the area corresponding to each LCZ block is used to classify each LCZ block on the windward side, thereby determining multiple windward side blocks and multiple leeward side blocks.

[0021] Optionally, the PM 2.5 The distance between the concentration movement test point and the road-based mobile source includes PM2.5 concentration. 2.5 Distance between concentration testing point and road mobile source along the prevailing wind direction, PM 2.5 The vertical road distance between the concentration testing point and the road-borne mobile source; based on the meteorological data, the PM2.5 concentration in each windward and leeward block is calculated. 2.5 The distance between the concentration movement test point and the road-based mobile source includes:

[0022] PM2.5 samples were collected from the windward and leeward sides of each of the aforementioned blocks.2.5 Geographic latitude and longitude coordinates of concentration movement test points and representative geographic latitude and longitude coordinates of road movement sources;

[0023] The spherical distance calculation formula is used based on the PM. 2.5 Using the geographic latitude and longitude coordinates of the concentration movement test point and the representative geographic latitude and longitude coordinates of the road movement source, PM2.5 is calculated. 2.5 The baseline Euclidean distance between the geographic latitude and longitude coordinates of the concentration movement test point and the representative geographic latitude and longitude coordinates of the road movement source;

[0024] Based on the meteorological data, the standard azimuth of the prevailing wind direction is determined, and the PM is calculated. 2.5 The azimuth angle of the line connecting the concentration movement test point relative to the road movement source;

[0025] Calculate the angle between the azimuth of the connecting line and the standard azimuth of the prevailing wind direction, and multiply the reference Euclidean distance by the cosine of the angle between the connecting line and the standard azimuth of the prevailing wind direction to obtain the PM. 2.5 Distance between the concentration testing point and the road-borne mobile source along the prevailing wind direction;

[0026] Calculate the PM 2.5 The angle between the line connecting the concentration movement test point and the road movement source and the direction perpendicular to the road, and the reference Euclidean distance is used to calculate the PM2.5 concentration. 2.5 The PM2.5 concentration is obtained by multiplying the cosine of the angle between the line connecting the concentration movement test point and the road movement source and the direction perpendicular to the road. 2.5 The vertical road distance between the concentration test point and the road-moving source.

[0027] Optionally, the windward-leeward level attenuation region includes a windward-side level attenuation region and a leeward-side level attenuation region; the windward-leeward initial attenuation prediction model includes a windward initial attenuation prediction model and a leeward initial attenuation prediction model; the morphological attenuation attributes and PM of the target blocks in each of the windward-leeward level attenuation regions are used. 2.5 The distance between the concentration movement test point and the road movement source is used to perform multi-level attenuation region iterative fitting on the initial attenuation prediction model on the windward and leeward sides to determine the target attenuation prediction model on the windward and leeward sides, including:

[0028] Based on the morphological attenuation properties of the target blocks in the N-level attenuation region within each of the aforementioned windward attenuation levels and PM 2.5 The distance between the concentration test point and the road mobile source along the prevailing wind direction is used to fit the Nth group of windward LCZ parameters, where N is the number of the windward attenuation level region, and N≥1;

[0029] The windward LCZ parameters of the Nth group are set as known quantities, and combined with the morphological attenuation properties of the target blocks in the N+1 level attenuation region of each windward attenuation region and PM 2.5 The distance between the concentration test point and the road mobile source along the prevailing wind direction is used to fit the N+1th group of windward LCZ parameters.

[0030] If the currently fitted windward level attenuation region is the last windward level attenuation region that was not fitted, and the windward LCZ parameter corresponding to the windward level attenuation region is fitted, the windward LCZ parameter obtained from this fitting is used as the windward attenuation factor matrix.

[0031] The initial attenuation prediction model on the windward side is updated using the windward attenuation factor matrix, and the target attenuation prediction model on the windward side is output.

[0032] Based on the morphological attenuation properties of the target blocks in the M-level attenuation region within each of the aforementioned leeward attenuation levels and PM 2.5 The vertical road distance between the concentration test point and the road mobile source is used to fit the Mth group of leeward LCZ parameters, where M is the sequence number of the leeward attenuation level region, and M≥1;

[0033] The leeward LCZ parameters of the Mth group are set as known quantities, and combined with the morphological attenuation properties of the target blocks in the M+1 level attenuation region of each leeward level attenuation region and PM 2.5 The vertical road distance between the concentration test point and the road-borne mobile source is used to fit the leeward LCZ parameters of the (M+1)th group.

[0034] If the currently fitted leeward-side level attenuation region is the last leeward-side level attenuation region that was not fitted, and after the leeward-side LCZ parameters corresponding to the leeward-side level attenuation region are fitted, the leeward-side LCZ parameters obtained from this fitting are used as the leeward-side attenuation factor matrix.

[0035] The leeward attenuation factor matrix is ​​used to update the initial leeward attenuation prediction model, and the target leeward attenuation prediction model is output.

[0036] Specifically, the initial attenuation prediction model on the windward side is as follows:

[0037] ;

[0038] In the formula, PM on the windward side 2.5 Concentration prediction; The attenuation factor for the target block with LCZ type i is preset in the windward attenuation region; , , , , , The LCZ parameters on the windward side are to be fitted to the initial attenuation prediction model on the windward side. The number of preset LCZ types that the pollutants pass through on the windward diffusion path in the diffusion path; The sky angle coefficient factor is preset for the target block with LCZ type i in the windward attenuation region; The aspect ratio of the street canyon of the target block with LCZ type i preset in the windward attenuation region; The roughness element height of the target block with LCZ type i is preset in the windward attenuation region; Preset the building density of the target block with LCZ type i in the windward attenuation region; The impermeable surface ratio of the target block with LCZ type i is preset in the windward attenuation region; The permeable surface ratio of the target block with LCZ type i is preset in the windward attenuation region; For PM 2.5 Distance between the concentration testing point and the road-borne mobile source along the prevailing wind direction; Preset the emission intensity of traffic mobile sources with LCZ type i in the windward attenuation zone;

[0039] The initial attenuation prediction model on the leeward side is specifically as follows:

[0040] ;

[0041] In the formula, leeward side PM 2.5 Concentration prediction; The number of preset LCZ types that the pollutant passes through on the leeward side of the diffusion path; The attenuation factor for the target block with LCZ type i is preset in the leeward attenuation region; , , , , , The leeward LCZ parameters to be fitted in the initial attenuation prediction model on the leeward side; Set the sky angle coefficient factor for the target block with LCZ type i in the leeward attenuation region; The aspect ratio of the street canyon of the target block with LCZ type i preset in the leeward level attenuation region; The roughness element height of the target block with LCZ type i is preset in the leeward level attenuation region; Preset the building density of the target block with LCZ type i in the leeward attenuation region; The impermeable surface ratio of the target block with LCZ type i is preset in the leeward level attenuation region; The permeable surface ratio of the target block with LCZ type i is preset in the leeward-side attenuation region; For PM 2.5 The vertical road distance between the concentration testing point and the road-borne mobile source; The emission intensity of traffic mobile sources with LCZ type i is preset in the leeward-side level attenuation region.

[0042] Optionally, it also includes:

[0043] Using the ordinary Kriging method based on the windward side PM 2.5 Concentration prediction and the leeward side PM 2.5 Spatial interpolation of concentration prediction values ​​is performed to output PM2.5. 2.5 Concentration field;

[0044] Spatial autocorrelation analysis and hotspot analysis were used to analyze the PM. 2.5 The concentration field is analyzed, and the spatial autocorrelation analysis results and hotspot analysis results are output.

[0045] The kernel density estimation method was used to analyze the PM on the windward side. 2.5 Concentration prediction and the leeward side PM 2.5 The predicted concentration values ​​are analyzed, and the results of the nuclear density analysis on the windward and leeward sides are output.

[0046] The second aspect of this invention provides a PM based on the theory of street morphological attenuation. 2.5 Concentration prediction device, including:

[0047] The acquisition module is used to acquire urban morphology information and meteorological data of the target area, and output the LCZ classification map of the target area and the morphological attenuation attributes of each LCZ block in the LCZ classification map according to multiple preset LCZ types and the urban morphology information.

[0048] The output module is used to classify multiple LCZ blocks in the LCZ classification map based on the meteorological data and output multiple windward and leeward blocks.

[0049] The calculation module is used to calculate the PM in each of the windward and leeward blocks. 2.5 The distance between the concentration movement test point and the road-based mobile source;

[0050] The segmentation module is used to determine the pollutant diffusion path based on the meteorological data, and to take the windward and leeward blocks within the pollutant diffusion path as target blocks, and to divide the multiple target blocks into attenuation regions to obtain multiple windward and leeward level attenuation regions.

[0051] The iterative module is used to construct the initial attenuation prediction model on the windward and leeward sides, and adopts the morphological attenuation attributes and PM of the target blocks in each of the windward and leeward attenuation level regions. 2.5 The distance between the concentration movement test point and the road movement source is used to perform multi-level attenuation region iterative fitting on the initial attenuation prediction model on the windward and leeward sides to determine the target attenuation prediction model on the windward and leeward sides.

[0052] The prediction module is used to predict PM levels on the windward and leeward sides of the target area using the windward and leeward side target attenuation prediction model. 2.5 Concentration prediction, outputting the PM2.5 concentration on the windward side of the target area. 2.5 Predicted concentration values ​​and leeward PM 2.5 Predicted concentration values.

[0053] A computer device provided in a third aspect of the present invention includes a memory and a processor. The memory stores a computer program, which, when executed by the processor, causes the processor to perform PM based on the street morphology attenuation theory as described in any of the preceding claims. 2.5 Steps in concentration prediction methods.

[0054] The fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed, implements PM based on the street morphological attenuation theory as described in any of the preceding claims. 2.5 Steps in concentration prediction methods.

[0055] A fifth aspect of the present invention provides a computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions, wherein, when the program instructions are executed by a computer, the computer performs PM based on the street morphology attenuation theory as described in any of the preceding claims. 2.5 Steps in concentration prediction methods.

[0056] As can be seen from the above technical solutions, the present invention has the following advantages:

[0057] The above-mentioned technical solution of the present invention provides a PM based on the theory of street morphological attenuation. 2.5The concentration prediction method acquires urban morphology information and meteorological data for the target area. Based on multiple preset LCZ types and urban morphology information, it outputs an LCZ classification map of the target area and the morphological attenuation attributes of each LCZ block in the LCZ classification map. Based on meteorological data, it classifies multiple LCZ blocks in the LCZ classification map, outputting multiple windward and leeward side blocks. Based on meteorological data, it calculates the PM2.5 concentration of each windward and leeward side block. 2.5 The distance between the concentration movement test point and the road-borne mobile source was determined; based on meteorological data, the pollutant diffusion path was determined, and the windward and leeward blocks within the pollutant diffusion path were designated as target blocks. Multiple target blocks were then divided into attenuation zones, resulting in multiple windward and leeward attenuation zones. An initial attenuation prediction model for the windward and leeward sides was constructed, and the morphological attenuation attributes of the target blocks within each windward and leeward attenuation zone and PM2.5 levels were used. 2.5 The distance between the concentration movement test point and the road-based mobile source was used to perform multi-level attenuation region iterative fitting on the initial attenuation prediction model on the windward and leeward sides to determine the target attenuation prediction model on both sides. The target area was then analyzed for PM2.5 concentrations on both the windward and leeward sides using the windward and leeward side target attenuation prediction model. 2.5 Concentration prediction, outputting the PM2.5 concentration on the windward side of the target area. 2.5 Predicted concentration values ​​and leeward PM 2.5 Concentration prediction; Based on the above scheme, this invention deeply integrates the LCZ classification system with the side-by-side zoning analysis logic, and accurately quantifies the impact of street blocks on PM through morphological attenuation attributes. 2.5 The diffusion attenuation mechanism fills the technical gaps in traditional models that do not associate LCZ attributes and do not quantify the attenuation effect of blocks. At the same time, multi-level iterative fitting does not rely on supercomputing resources and complex parameter inputs, which simplifies the model's computational cost and allows the model to fully characterize the pollution diffusion patterns under different spatial dimensions through refined fitting of sides (windward / leeward) and levels (attenuation areas), significantly improving prediction accuracy. Attached Figure Description

[0058] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0059] Figure 1 This invention provides a PM based on the street morphology attenuation theory, as shown in Embodiment 1. 2.5 Flowchart of the concentration prediction method steps;

[0060] Figure 2This is a schematic diagram of a preset LCZ type provided in Embodiment 1 of the present invention;

[0061] Figure 3 This is a spatial distribution map of morphological attenuation properties of the Nanda Expressway area in Guangzhou City, as the target area, provided in Embodiment 1 of the present invention.

[0062] Figure 4 This is an LCZ classification map of the Guangzhou South Trunk Line area provided in Embodiment 1 of the present invention;

[0063] Figure 5 This is a schematic diagram of the overall framework of the particle concentration prediction model provided in Embodiment 1 of the present invention;

[0064] Figure 6 This is a schematic diagram of the framework for iterative calculation of the inherent attenuation factor provided in Embodiment 1 of the present invention;

[0065] Figure 7 This is a schematic diagram of the framework for iterative calculation of the additional attenuation factor provided in Embodiment 1 of the present invention;

[0066] Figure 8 This invention provides a PM based on the street morphology attenuation theory, as shown in Embodiment 1. 2.5 A flowchart illustrating the concentration prediction method;

[0067] Figure 9 The windward PM output is provided by the target attenuation prediction model applied to the windward and leeward sides in Embodiment 2 of the present invention. 2.5 Predicted concentration values ​​and leeward PM 2.5 A flowchart illustrating the process of analyzing predicted concentration values;

[0068] Figure 10 This invention provides a PM based on the street morphology attenuation theory, as shown in Embodiment 3. 2.5 Block diagram of the concentration prediction device. Detailed Implementation

[0069] This invention provides a PM based on the theory of street morphological attenuation. 2.5 Concentration prediction methods and devices for solving traditional PM2.5 concentration problems. 2.5 The prediction model relies on supercomputing and complex parameters, and lacks quantification of the street decay mechanism and LCZ attribute correlation, thus failing to fully characterize the diffusion law, ultimately resulting in limited prediction accuracy. It is suitable for high-density cities in hot and humid climate zones.

[0070] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, 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. It should be noted that in the optional embodiments of the present invention, the object information and other related data involved require the permission or consent of the object when the embodiments of the present invention are applied to specific products or technologies, and the collection, use, and processing of related data must comply with the relevant laws, regulations, and standards of the relevant countries and regions. That is to say, if the embodiments of the present invention involve data related to the object, it needs to be obtained with the authorization and consent of the object, the authorization and consent of the relevant departments, and in compliance with the relevant laws, regulations, and standards of the country and region. If personal information is involved in the embodiments, the acquisition of all personal information requires the consent of the individual. If sensitive information is involved, the separate consent of the information subject is required, and the embodiments also need to be implemented with the authorization and consent of the object.

[0071] Please see Figure 1 , Figure 1 This invention provides a PM based on the street morphology attenuation theory, as shown in Embodiment 1. 2.5 Flowchart of the concentration prediction method.

[0072] This invention provides a PM based on the theory of street morphological attenuation. 2.5 Concentration prediction methods include:

[0073] Step 101: Obtain urban morphology information and meteorological data of the target area, and output the LCZ classification map of the target area and the morphological attenuation attributes of each LCZ block in the LCZ classification map based on multiple preset LCZ types and urban morphology information.

[0074] The target area refers to the area where PM is planned to be carried out in this invention. 2.5 Concentration prediction and attenuation zone delineation typically refer to specific urban spatial areas affected by PM2.5 transported by road traffic sources (such as a main road). 2.5 The core area of ​​the diffusion impact can be determined through administrative planning or pollution impact range calculation, such as the Nanda Expressway area in Guangzhou.

[0075] Urban morphology information includes road network data, remote sensing images, and surface parameter data. Road network data is structured data reflecting the spatial distribution, type attributes, and functional characteristics of the road network within the target area, including road geometry, road type classification, road network topology, and related information (such as the number of lanes and speed limits). Remote sensing images are surface spatial image data of the target area acquired by sensors mounted on platforms such as satellites, drones, and aerial remote sensing. Surface parameter data is a quantitative data set that quantifies the land cover type, underlying surface physical characteristics, and ecological attributes within the target area, primarily covering three types of parameters: first, land cover parameters (percentage of impervious surface, vegetation coverage, and water area percentage, reflecting the surface's ability to hinder pollutant diffusion and sedimentation); second, building morphology parameters (building density, average building height, and building spacing, directly related to LCZ type determination); and third, underlying surface physical parameters (surface albedo and roughness length, affecting local microclimate and airflow, indirectly impacting PM2.5). 2.5 diffusion).

[0076] Meteorological data supports the analysis of pollutant diffusion paths and the division of sides (windward / leeward). It mainly includes the prevailing wind direction, average wind speed, wind direction frequency, and atmospheric stability of the target area. It contains both historical statistical data (used to determine long-term diffusion patterns) and real-time data (used for dynamic prediction and adjustment).

[0077] The preset LCZ types include: Compact High-Rise Building Zone (LCZ1), Compact Mid-Rise Building Zone (LCZ2), Compact Low-Rise Building Zone (LCZ3), Open High-Rise Building Zone (LCZ4), Open Mid-Rise Building Zone (LCZ5), Open Low-Rise Building Zone (LCZ6), Simple Low-Rise Building Zone (LCZ7), Large Low-Rise Building Zone (LCZ8), Ultra-Open Low-Rise Building Zone (LCZ9), Heavy Industrial Zone (LCZ10), Dense Tree Zone (LCZA), Sparse Tree Zone (LCZB), Shrub Zone (LCZC), Low Vegetation Zone (LCZD), Hardened Ground Zone (LCZE), Bare Sand Zone (LCZF), and Water Zone (LCZG). Figure 2 As shown.

[0078] Morphological decay properties for quantized LCZ blocks to PM 2.5The morphological parameters of diffusion attenuation effects include the Sky View Factor (SVF), the Aspect Ratio (AR) of street canyons, the Height of Roughness Elements (HRE), the Building Surface Factor (BSF), the Impervious Surface Ratio (ISF), and the Pervious Surface Ratio (PSF). Specifically, the Sky View Factor (SVF) is the ratio of the number of visible sky hemispheres to the number of unobstructed sky hemispheres; the Aspect Ratio (AR) of street canyons is the average aspect ratio of street canyons; the Height of Roughness Elements (HRE) is the geometric mean of building height and tree / vegetation height; the Building Surface Factor (BSF) is the ratio of the area of ​​building plan to the total planar area; the Impervious Surface Ratio (ISF) is the ratio of the area of ​​impervious surfaces to the total planar area; and the Pervious Surface Ratio (PSF) is the ratio of the area of ​​permeable surfaces to the total planar area. For example, ... Figure 3 As shown, Figure 3 This paper presents a spatial distribution map of morphological attenuation properties in the Nanda Expressway area of ​​Guangzhou.

[0079] It should be noted that the process begins by retrieving road network data and high-resolution remote sensing images of the target area through a GIS platform, combined with surface parameters collected through on-site surveys (such as building density and vegetation cover), to form urban morphology information. Simultaneously, historical and real-time meteorological data, such as prevailing wind direction and average wind speed, are obtained from regional meteorological stations. Then, based on preset LCZ types (such as compact high-rise, open low-rise, and dense vegetation areas), spatial matching of the urban morphology information is performed. ArcGIS spatial analysis tools are used to divide the target area into grids and label block types, generating an LCZ classification map. Finally, for each LCZ block in the map, morphological attenuation attributes are extracted and calculated using remote sensing inversion (calculating SVF), on-site measurements (obtaining building aspect ratio AR), and parametric statistics (calculating BSF and ISF), forming a corresponding dataset of "LCZ block - morphological attenuation attributes".

[0080] Specifically, step 101 may include the following sub-steps:

[0081] S11. Preprocess the road network data and output the urban road network of the target area;

[0082] S12. Use the urban road network of the target area as the LCZ block segmentation boundary, and based on the LCZ block segmentation boundary, use GIS spatial analysis methods to segment the target area according to remote sensing data, and output multiple initial LCZ blocks.

[0083] S13. Extract the morphological decay attributes of each initial LCZ block, and classify each initial LCZ block according to the morphological decay attributes of each initial LCZ block based on multiple preset LCZ types, and output multiple LCZ blocks with preset LCZ type labels.

[0084] S14. Verify each LCZ block with a preset LCZ type label, generate multiple LCZ blocks, and use multiple LCZ blocks to construct an LCZ classification map of the target area.

[0085] It should be noted that when constructing the LCZ classification map, the road network data of the target area is first preprocessed (e.g., correcting data errors and unifying the coordinate system) to output an accurate urban road network. This urban road network is then used as the dividing boundary for LCZ blocks. Combined with remote sensing data, the target area is spatially segmented using GIS spatial analysis methods (e.g., overlay analysis and grid division) to obtain multiple initial LCZ blocks. Subsequently, the morphological attenuation attributes (e.g., SVF, AR, ISF) of each initial LCZ block are extracted. Based on the feature thresholds of multiple preset LCZ types (e.g., compact high-rise LCZ1 corresponds to a building height ≥15m and an impermeable surface ratio ≥70%), the morphological attenuation attributes of each initial LCZ block are matched and classified with the preset types, outputting multiple LCZ blocks with preset LCZ type labels. Finally, these labeled LCZ blocks are verified (e.g., on-site sampling to check if the block type matches the actual shape and correcting classification deviations) to generate multiple accurate LCZ blocks. These blocks are then integrated according to spatial location to ultimately construct an LCZ classification map covering the target area.

[0086] For example, such as Figure 4As shown, taking the Nanda Trunk Line area of ​​Guangzhou as the target area as an example, when constructing an LCZ map of the Nanda Trunk Line area of ​​Guangzhou, a Geographic Information System (GIS) combined with Google Earth is used to focus on the urban streets on both sides of the road in the area for LCZ division. The core steps are consistent with the previous logic: First, the LCZ dividing boundary is determined with road network data as the core. Since blocks are the basic building blocks of the city structure, and the roads in the Nanda Trunk Line area are crisscrossed, the blocks they form, although the building uses are complex, can naturally present a square or rectangular shape based on the road network division. Therefore, the road network data of the area is preprocessed (such as calibrating coordinates and correcting data errors), and then the preprocessed urban road network is selected as the dividing boundary of the LCZ blocks. Second, data collection and attribute calculation are completed. Surface parameter data (such as building density, vegetation cover, and impervious surface ratio) are collected in each LCZ division unit. Based on these basic parameters, GIS tools are used to calculate the LCZ morphological attenuation attributes (such as SVF, AR, ISF, etc.) of each division unit, and the PM of the roads in the area is integrated simultaneously. 2.5 The monitoring data (including data from fixed stations and data collected by mobile sensors) and meteorological data (including wind speed, wind direction, and air pressure gradient) provide complete data support for subsequent LCZ block classification and accuracy verification.

[0087] Step 102: Based on meteorological data, classify multiple LCZ blocks in the LCZ classification map and output multiple windward and leeward blocks.

[0088] The windward and leeward blocks include windward and leeward blocks; among them, the windward block refers to the LCZ block located in the direction of "road mobile source → prevailing wind direction", which is located in the PM 2.5 Along the main diffusion path, PM2.5 is most directly affected by emissions from roadside mobile sources. 2.5 The core area for concentration decay simulation; leeward side block: refers to the LCZ block located "perpendicular to the road direction, on both sides of the road mobile source", in the PM 2.5 Along the lateral diffusion path, the impact of emissions from roadside mobile sources is weaker than on the windward side, and its PM2.5 concentration is lower. 2.5 The concentration decay pattern differs significantly from that on the windward side. The perpendicular road direction refers to the horizontal direction perpendicular to the centerline of the corresponding road (or the road segment upon which the road mobile source relies) within the target area, including two opposite directions extending along this vertical direction to both sides of the road. It is a characterizing factor for PM2.5 concentrations. 2.5 The core benchmark is the spatial relationship between concentration test points and road mobile sources in the lateral dimension of the road.

[0089] Road traffic mobile sources: refers to PM emissions within the target area 2.5Major road pollution sources are typically defined using the road centerline as a spatial reference (e.g., the centerline of the South Trunk Road), which serves as the spatial reference point for dividing windward and leeward zones. PM2.5 2.5 The diffusion radiates to the surrounding blocks based on this benchmark.

[0090] Specifically, step 102 may include the following sub-steps:

[0091] S21. Use meteorological data to divide the target area into wind-side zones, and output the global zone range on the windward and leeward sides of the target area.

[0092] S22. Based on the global partition range on the windward side and the global partition range on the leeward side, the area corresponding to each LCZ block is used to classify each LCZ block on the windward side, and multiple windward side blocks and multiple leeward side blocks are determined.

[0093] It should be noted that core parameters are extracted from meteorological data (such as the long-term prevailing wind direction and average wind speed of the target area), and combined with identified road traffic sources (such as the centerline of the South Trunk Road in Guangzhou), wind-side zoning is carried out using GIS spatial analysis tools (such as direction analysis and buffer zone analysis): using the road centerline as a reference, extending outward along the prevailing wind direction to PM... 2.5 The main diffusion impact range (e.g., 500 meters) is defined as a continuous spatial area as the windward global zoning range; simultaneously, an equal impact range is extended along both sides perpendicular to the road direction, and a continuous spatial area is defined as the leeward global zoning range, thus completing the windward zoning of the target area and outputting two types of global ranges. Subsequently, each LCZ block in the LCZ classification map is spatially overlaid with the windward and leeward global zoning ranges, respectively. The area percentage of a single LCZ block falling within the two types of global ranges is calculated. If the area percentage of a certain LCZ block falling within the windward global zoning range is ≥50%, it is determined to be a windward block; if the area percentage falling within the leeward global zoning range is ≥50%, it is determined to be a leeward block. Finally, multiple windward blocks and multiple leeward blocks are determined.

[0094] The leeward side global zone is defined as follows: based on road traffic movement sources (such as the road centerline), extending to both sides along the direction perpendicular to the road to PM. 2.5The continuous spatial area formed by the secondary diffusion influence range is the global reference boundary for determining whether an LCZ block is a "leeward block". The area corresponding to an LCZ block refers to the total area of ​​a single LCZ block in the LCZ classification map, and the portion of that block falling within the "windward global zoning range" and "leeward global zoning range" respectively. This is the core quantitative basis for determining the wind-side attribute of the block. The dominant wind direction refers to the wind direction with the highest frequency in the target area over a certain period (usually one year or more), representing the main direction of atmospheric flow in the area and directly determining PM2.5 concentration. 2.5 The main path of diffusion from road-borne mobile sources is the fundamental criterion for distinguishing between the windward and leeward sides.

[0095] Step 103: Based on meteorological data, calculate the PM2.5 concentration in each windward and leeward block. 2.5 The distance between the concentration movement test point and the road-based mobile source.

[0096] PM 2.5 The distance between the concentration movement test point and the road-based mobile source includes PM2.5 concentration. 2.5 Distance between concentration testing point and road mobile source along the prevailing wind direction, PM 2.5 The vertical road distance between the concentration test point and the road-moving source.

[0097] It should be noted that, based on the prevailing wind direction, the wind direction under study is divided into the prevailing wind (windward side) and the non-prevailing wind (leeward side). PM2.5 under the influence of the prevailing wind... 2.5 The relationship between particulate matter concentration and distance is: ( Road PM 2.5 Monitoring data; PM 2.5 (Distance between concentration testing point and road-based mobile source along the prevailing wind direction); PM2.5 under the influence of non-prevailing winds 2.5 The relationship between particulate matter concentration and distance is: ( PM 2.5 (Vertical road distance between the concentration test point and the road-borne mobile source).

[0098] Specifically, step 103 may include the following sub-steps:

[0099] S31. Collect PM2.5 concentrations in each windward and leeward side area. 2.5 Geographic latitude and longitude coordinates of concentration movement test points and representative geographic latitude and longitude coordinates of road movement sources;

[0100] S32, Using the spherical distance calculation formula based on PM 2.5 The geographic latitude and longitude coordinates of the concentration movement test points and the representative geographic latitude and longitude coordinates of road movement sources are used to calculate PM2.5.2.5 The baseline Euclidean distance between the geographic latitude and longitude coordinates of the concentration movement test point and the representative geographic latitude and longitude coordinates of the road movement source;

[0101] S33. Determine the standard azimuth of the prevailing wind direction based on meteorological data, and calculate PM2.5. 2.5 The azimuth angle of the line connecting the concentration movement test point to the road-moving source;

[0102] S34. Calculate the angle between the azimuth of the connecting line and the standard azimuth of the prevailing wind direction, and multiply the reference Euclidean distance by the cosine of the angle between the connecting line and the standard azimuth of the prevailing wind direction to obtain the PM. 2.5 Distance between the concentration testing point and the road-borne mobile source along the prevailing wind direction;

[0103] S35, Calculate PM 2.5 The angle between the line connecting the concentration movement test point and the road movement source and the direction perpendicular to the road is calculated, and the reference Euclidean distance is used to calculate PM2.5. 2.5 The PM2.5 concentration is obtained by multiplying the cosine of the angle between the line connecting the concentration movement test point and the road-based mobile source and the direction perpendicular to the road. 2.5 The vertical road distance between the concentration test point and the road-moving source.

[0104] It should be noted that the calculation of PM 2.5 The distances between concentration testing points and road-borne mobile sources along the prevailing wind direction and perpendicular to the road direction need to be determined based on coordinate acquisition and baseline distance calculation, combined with corresponding directional baselines to complete projection decomposition. The specific process is as follows: First, perform the coordinate acquisition step to collect PM2.5 concentration data from each windward and leeward side block. 2.5The geographic latitude and longitude coordinates of the concentration movement test point, and the representative geographic latitude and longitude coordinates of the road movement source (preferably the centerline coordinates of the road closest to the movement test point, or the center point coordinates of the road segment adjacent to the target block to which the test point belongs); then, based on the latitude and longitude coordinates of the above two points, the spherical distance calculation formula is used to calculate the reference Euclidean distance between the two points, which serves as the basis data for subsequent directional distance decomposition; for the distance along the prevailing wind direction, the long-term prevailing wind direction is first determined by meteorological data of the target area and converted into a standard azimuth (measured clockwise with true north as 0°), and then the distance relative to the test point is calculated. The azimuth angle of the line connecting the representative coordinates of the road moving source is calculated, and the angle between this line and the prevailing wind direction is obtained. The distance along the prevailing wind direction is obtained by multiplying the reference Euclidean distance by the cosine of this angle. For the distance perpendicular to the road direction, the azimuth angle of the road direction is first determined by fitting the coordinates of the road centerline, and then the reference azimuth angle perpendicular to the road direction (road azimuth ± 90°) is derived. Then, the angle between the line connecting the test point and the representative coordinates of the road moving source and the perpendicular road direction is calculated. The distance perpendicular to the road direction is obtained by multiplying the reference Euclidean distance by the cosine of this angle. Finally, the distance data in the corresponding directions are output.

[0105] Step 104: Based on meteorological data, determine the pollutant diffusion path, and take the windward and leeward blocks within the pollutant diffusion path as target blocks, and divide the multiple target blocks into attenuation areas to obtain multiple windward and leeward level attenuation areas.

[0106] Pollutant diffusion pathways refer to pathways constructed based on road traffic mobile sources, prevailing wind direction, and wind speed, covering PM2.5. 2.5 The core spatial range from emissions to diffusion is clearly divided into windward diffusion paths and leeward diffusion paths, which are used to accurately screen the core areas affected by pollution.

[0107] Target Block: Refers to the windward / leeward side block within the pollutant diffusion path. It is clearly divided into target windward block and target leeward block. It is the core object for subsequent attenuation area division and model fitting. It must meet the screening condition of "completely or mostly (area ratio ≥60%) within the corresponding diffusion path".

[0108] The windward and leeward attenuation zones include both windward and leeward attenuation zones. The windward attenuation zone is a zone divided based on the target windward block, graded according to the number of target windward blocks that pollutants sequentially cross from the road-moving source along the prevailing wind direction. Higher grades indicate greater distance from the pollution source and lower PM2.5 concentrations. 2.5The higher the concentration attenuation, the better. Leeward-side attenuation zone: Attenuation zones are defined based on the target leeward side blocks, graded according to the number of target leeward side blocks that pollutants sequentially traverse from the road-moving source along the perpendicular road direction. The grading logic is consistent with the windward side, used to distinguish PM at different distances on the leeward side. 2.5 Attenuation characteristics.

[0109] It should be noted that when determining pollutant diffusion paths and completing target block screening and attenuation area delineation based on meteorological data, the long-term prevailing wind direction and average wind speed of the target area are first extracted from the meteorological data. Starting from a clearly identified road traffic mobile source (such as the road centerline), the windward and leeward diffusion paths of pollutants are constructed respectively, extending along the prevailing wind direction to PM. 2.5 The linear area of ​​the primary diffusion influence range (set according to the average wind speed, e.g., 500 meters when the wind speed is 1-3 m / s) is the windward diffusion path; the area extending to the secondary diffusion influence range (e.g., 300 meters) along both sides perpendicular to the road direction is the leeward diffusion path. Together, they constitute a complete pollutant diffusion path. Subsequently, the two types of diffusion paths are spatially superimposed with the previously divided windward and leeward blocks, respectively. Windward blocks that are completely or mostly (area percentage ≥ 60%) within the windward diffusion path are selected as target windward blocks; leeward blocks that are completely or mostly within the leeward diffusion path are selected as target leeward blocks. Together, these two types of blocks constitute the target block. Finally, the target blocks are divided into attenuation zones based on "side + number of crossings": the target windward blocks are divided into first-level windward attenuation zones according to the order of pollutants moving outward from the road source along the prevailing wind direction, and the first target windward block that pollutants cross in sequence is divided into second-level windward attenuation zones, and so on; the target leeward blocks are divided into first-level leeward attenuation zones according to the order of pollutants moving outward from the road source perpendicular to the road direction, and the first target leeward block that pollutants cross in sequence is divided into second-level leeward attenuation zones, and so on. In the end, multiple windward attenuation zones and multiple leeward attenuation zones are obtained.

[0110] Step 105: Construct initial attenuation prediction models for the windward and leeward sides, and use the morphological attenuation attributes and PM of target blocks in the attenuation regions of each windward and leeward side level. 2.5 The distance between the concentration movement test point and the road movement source is used to perform multi-level attenuation region iterative fitting on the initial attenuation prediction model on the windward and leeward sides to determine the target attenuation prediction model on the windward and leeward sides.

[0111] The initial attenuation prediction model for the windward and leeward sides includes the initial attenuation prediction model for the windward side and the initial attenuation prediction model for the leeward side.

[0112] It should be noted that, as Figure 5 This is a schematic diagram of the overall framework of the particulate concentration prediction model. Since particulate matter is transported downwind under the influence of the prevailing wind, its concentration decay is affected by the intensity of turbulent diffusion. This invention defines a decay factor (…). It is a characteristic correction factor and attenuation factor for the pollutant diffusion process in different LCZ blocks. The value is determined by both inherent and additional attributes. Furthermore, traffic source PM 2.5 Primarily originating from vehicle exhaust emissions, pollutants, after being released into the atmosphere, are first affected by the background wind field, gradually diffusing and diluting towards roadside areas along the prevailing wind direction. During this process, due to significant differences in spatial characteristics among different LCZ (Limited Zone) blocks, the resistance effect on pollutant diffusion varies from block to block. These differences directly affect the concentration distribution and attenuation rate of pollutants on the windward and leeward sides.

[0113] Furthermore, after extracting the morphological attenuation attributes of each LCZ block, it is necessary to decouple the morphological attenuation attributes into intrinsic attributes and additional attributes. Among them, intrinsic attributes refer to the spatial characteristic attributes inherent to each LCZ type, such as roughness element height (HRE) and building density (BSF); additional attributes refer to the differences in specific spatial morphological parameters of each block unit within the same LCZ type, such as impermeable surface ratio (ISF) and permeable surface ratio (PSF).

[0114] Furthermore, the initial attenuation prediction model on the windward side is as follows:

[0115] ;

[0116] In the formula, PM on the windward side 2.5 The concentration prediction value represents the result of the model's concentration prediction, multiplied by the cumulative effect of pollutant concentration decay; The attenuation factor is the target block with a preset LCZ type of i in the windward attenuation zone. It represents the hindering or promoting effect of the preset LCZ type on pollutant diffusion. i=1,2,3,4,5,6,9,A,D represent the specific LCZ types of interest in the model (LCZ1-6, LCZ9, LCZ A, LCZ D). The specific LCZ types include compact high-rise building area LCZ1, compact mid-rise building area LCZ2, compact low-rise building area LCZ3, open high-rise building area LCZ4, open mid-rise building area LCZ5, open low-rise building area LCZ6, ultra-open low-rise building area LCZ9, dense tree area LCZA, and low vegetation area LCZD. , , , , , The LCZ parameters on the windward side are to be fitted to the initial attenuation prediction model on the windward side. The number of preset LCZ types that the pollutants pass through on the windward diffusion path in the diffusion path; The sky angle coefficient factor for the target block with LCZ type i is preset in the windward attenuation region, representing the proportion of the sky visible in the street canyon; The aspect ratio of the street canyon for the target block with LCZ type i in the windward attenuation region is the ratio of the average height of the buildings on both sides of the street to the width of the street. The roughness element height of the target block with LCZ type i is preset in the windward level attenuation region, representing the height of the average building or other surface roughness elements in the block; The building density of the target block with LCZ type i is preset in the windward attenuation region, which represents the proportion of the area occupied by buildings in the block; The impermeable surface ratio of the target block with LCZ type i preset in the windward attenuation region represents the proportion of impermeable areas such as hardened roads and squares. The permeable surface ratio of the target block with LCZ type i preset in the windward attenuation region, representing the proportion of permeable areas such as green space and lawn; For PM 2.5 Distance between the concentration testing point and the road-borne mobile source along the prevailing wind direction; Preset the emission intensity of traffic mobile sources with LCZ type i in the windward attenuation zone;

[0117] The initial attenuation prediction model on the leeward side is as follows:

[0118] ;

[0119] In the formula, leeward side PM 2.5 The concentration prediction value represents the result of the model's concentration prediction, multiplied by the cumulative effect of pollutant concentration decay; The number of preset LCZ types that the pollutant passes through on the leeward side of the diffusion path; The attenuation factor for the target block with LCZ type i is preset in the leeward attenuation region; , , , , , The leeward LCZ parameters to be fitted in the initial attenuation prediction model on the leeward side; Set the sky angle coefficient factor for the target block with LCZ type i in the leeward attenuation region; The aspect ratio of the street canyon of the target block with LCZ type i preset in the leeward level attenuation region; The roughness element height of the target block with LCZ type i is preset in the leeward level attenuation region; Preset the building density of the target block with LCZ type i in the leeward attenuation region; The impermeable surface ratio of the target block with LCZ type i is preset in the leeward level attenuation region; The permeable surface ratio of the target block with LCZ type i is preset in the leeward-side attenuation region; For PM 2.5 The vertical road distance between the concentration testing point and the road-borne mobile source; The emission intensity of traffic mobile sources with LCZ type i is preset in the leeward-side level attenuation region.

[0120] It is worth mentioning that this invention draws on the core idea of ​​the turbulent diffusion attenuation equation and combines it with a self-proposed morphological attenuation theory to construct initial attenuation prediction models for the windward and leeward sides, and defines an attenuation factor based on this theory. The attenuation factor, as a key parameter in the model formula, is determined through iterative fitting of the windward and leeward LCZ parameters to be fitted, along with the parameters to be fitted, through multi-level attenuation regions. Both are jointly determined by the inherent and additional properties of each region, and exhibit specificity with the difference in LCZ type. The attenuation factor and the aforementioned fitting parameters corresponding to different LCZ type blocks all possess unique characteristics. The turbulent diffusion attenuation equation refers to the equation describing the attenuation of pollutants in the atmosphere (such as PM2.5). 2.5 A mathematical model for PM2.5 diffusion due to atmospheric turbulence, and its concentration decay with diffusion distance, meteorological conditions, and underlying surface properties (such as LCZ block morphology), is designed to quantify the relationship between pollutant concentration and diffusion influencing factors. 2.5 Concentration prediction provides the core calculation basis.

[0121] Furthermore, step 105 may include the following sub-steps:

[0122] S51. Based on the morphological attenuation properties of the target blocks in the N-level attenuation region within each windward attenuation level and PM... 2.5 The distance between the concentration test point and the road mobile source along the prevailing wind direction is used to fit the Nth group of windward LCZ parameters, where N is the number of the windward attenuation level region, and N≥1;

[0123] S52. Set the LCZ parameters of the Nth group on the windward side as known quantities, and combine them with the morphological attenuation properties of the target block in the N+1 level attenuation region of each windward attenuation level and PM. 2.5The distance between the concentration test point and the road mobile source along the prevailing wind direction is used to fit the N+1th group of windward LCZ parameters.

[0124] S53. If the currently fitted windward level attenuation region is the last windward level attenuation region that has not been fitted, and after the windward LCZ parameter fitting corresponding to the windward level attenuation region is completed, the windward LCZ parameter obtained from this fitting is used as the windward attenuation factor matrix.

[0125] S54. Update the initial attenuation prediction model on the windward side using the windward attenuation factor matrix, and output the target attenuation prediction model on the windward side.

[0126] S55, based on the morphological attenuation properties of the target blocks in the M-level attenuation region within each leeward attenuation region and PM 2.5 The vertical road distance between the concentration test point and the road mobile source is used to fit the Mth group of leeward LCZ parameters, where M is the sequence number of the leeward attenuation level region, and M≥1;

[0127] S56. Set the leeward LCZ parameters of the Mth group as known quantities, and combine them with the morphological attenuation properties of the target block in the M+1 level attenuation region of each leeward attenuation region and PM. 2.5 The vertical road distance between the concentration test point and the road-borne mobile source is used to fit the leeward LCZ parameters of the (M+1)th group.

[0128] S57. If the currently fitted leeward-side level attenuation region is the last leeward-side level attenuation region that has not been fitted, and after the leeward-side LCZ parameter fitting corresponding to the leeward-side level attenuation region is completed, the leeward-side LCZ parameter obtained from this fitting is used as the leeward-side attenuation factor matrix.

[0129] S58. Update the initial attenuation prediction model on the leeward side using the leeward side attenuation factor matrix, and output the target attenuation prediction model on the leeward side.

[0130] It should be noted that, as Figure 6 Figure 7 is a schematic diagram of the framework for iterative calculation of the attenuation factor. Taking the windward attenuation zone as an example, it presents a significant multi-level attenuation zone along the wind direction. Based on the relative location of traffic sources and the number of land parcels traversed during pollutant propagation, it is clearly divided into Level 1, Level 2, Level 3, up to Level N attenuation zones. Each level of zone represents the PM2.5 concentration. 2.5 The number of LCZ blocks (i.e. target blocks) that the particle passes through in sequence during propagation.

[0131] (a) Select monitoring data from the primary attenuation area (single-plot diffusion), i.e., morphological attenuation attributes and PM2.5. 2.5The distance along the prevailing wind direction between the concentration test point and the road-borne mobile source is used to fit the attenuation factor of the first group of LCZ types.

[0132] (b) Select monitoring data of the secondary attenuation area (diffusion between two plots), use the results of step (a) as known quantities, fit the second set of LCZ types, and dynamically update the parameters of repeated LCZ types;

[0133] (c) Select monitoring data of the three-level attenuation area (diffusion of three plots), use the updated parameters in step (b) as known quantities, fit the third group of LCZ types, and update the repeating parameters;

[0134] (d) Select monitoring data of the N-level attenuation area (N-block diffusion), use the updated parameters in step (c) as known quantities, fit the Nth group of LCZ types, until all groups are completed, and output the final attenuation factor matrix, i.e. the windward attenuation factor matrix.

[0135] When the LCZ type appears repeatedly in two adjacent iterations, the fitted value of the later level completely replaces the historical value of the previous level; the parameters of the non-repeating LCZ type are locked in the first level when they appear, and are passed as constants in subsequent levels.

[0136] Specifically, for each level of attenuation zone on the windward side, starting from the N-level attenuation zone with sequence number N≥1, core data of all target windward blocks within this N-level attenuation zone are collected, including the morphological attenuation attributes of each target block (such as sky visibility factor SVF, building aspect ratio AR, impermeable surface ratio ISF, etc.) and the PM within each target block calculated in the previous stage. 2.5 The distances between concentration testing points and road-borne mobile sources along the prevailing wind direction were standardized (e.g., uniform units, outlier removal). Numerical fitting methods (e.g., least squares, multiple linear regression) were then used to fit and calculate the data, yielding the Nth set of windward LCZ parameters corresponding to the Nth-level attenuation region. This parameter set accurately reflects the influence of LCZ attributes and distance on PM2.5 levels within the Nth-level region. 2.5The comprehensive influence of concentration was investigated. Next, the fitted Nth group of windward LCZ parameters were set as known fixed values. Then, the same core data (morphological attenuation attributes, distance along the prevailing wind direction) of all target windward blocks within the N+1 level attenuation region were collected. After standardization, the LCZ types of the target blocks within this N+1 level attenuation region were analyzed, distinguishing between LCZ types that overlapped with the N level attenuation region and newly added LCZ types. For the newly added LCZ types at the N+1 level, their corresponding core data were substituted into the fitting model as unknowns. For each LCZ type, a separate fitting is performed. For LCZ types that repeat at level N+1 and level N, the fitting is performed again by combining the already fitted parameters at level N with the measured data of this type in the area of ​​level N+1. The parameter values ​​are updated to adapt to the attenuation characteristics of the area of ​​level N+1. After processing the parameters of the two types of LCZ, the N+1 group of windward side LCZ parameters is obtained. This achieves the iterative process of "the parameters of the previous level (level N) supporting the fitting of the parameters of the next level (level N+1), and the parameters of level N+1 specifically handling the newly added and repeated types". Then, the above iterative fitting steps are repeated continuously (such as fitting level N+1 LCZ). When dealing with attenuation regions at levels 2 and later, the LCZ type differences from the preceding N and N+1 attenuation regions are first identified. For newly added types, unknown quantity fitting is performed, and repeated types are refitted and updated to obtain the corresponding level's lateral LCZ parameters. This continues until the current windward attenuation region to be fitted is the last unfitted region among all windward attenuation regions. After fitting the windward LCZ parameters corresponding to this last attenuation region, all previously fitted windward LCZ parameters (Nth group, N+1th group...last group) are sequentially arranged and integrated according to the attenuation region number to construct a windward attenuation factor matrix containing the characteristics of each attenuation region level. This matrix can completely cover the parameter patterns of "new and repeated LCZ types" under different attenuation levels on the windward side. Finally, the constructed windward attenuation factor matrix is ​​substituted into the previously built initial windward attenuation prediction model, replacing the original general parameter terms within the model. The core calculation coefficients and attenuation law formulas of the model are adjusted and calibrated to ensure the model can accurately match PMs of different LCZ types in each attenuation region on the windward side. 2.5 After analyzing the concentration change characteristics and updating the model, the final output is a prediction model for the attenuation of windward targets that can be used for actual predictions. Similarly, a prediction model for the attenuation of leeward targets can be obtained.

[0137] Step 106: Analyze the PM levels of the target area on the windward and leeward sides using the windward and leeward side target attenuation prediction model. 2.5 Concentration prediction, outputting the PM2.5 concentration on the windward side of the target area. 2.5 Predicted concentration values ​​and leeward PM 2.5 Predicted concentration values.

[0138] It should be noted that the target area's PM levels on the windward and leeward sides are predicted using the target attenuation prediction model on both sides. 2.5 When predicting concentration, the core input data corresponding to the location to be predicted (spatial location) within the target area is first collected, including the morphological attenuation properties (such as SVF, AR, ISF, etc.) of the LCZ block to which the location belongs, and the PM at that location. 2.5 The distance between the concentration testing point and the road-borne mobile source is calculated (the distance along the prevailing wind direction for the windward prediction location, and the distance perpendicular to the road direction for the leeward prediction location). Then, the input data from the windward prediction location is substituted into the windward target attenuation prediction model. The model will then use the previously constructed windward attenuation factor matrix to calculate the PM2.5 concentration at that location. 2.5 Concentration; simultaneously, the input data of the leeward side to be predicted location is substituted into the leeward side target attenuation prediction model, and finally the PM concentrations of all windward side to be predicted locations within the target area are output. 2.5 Predicted concentration values ​​(which can generate a spatial distribution map or numerical table of concentrations on the windward side), and PM2.5 concentrations at all locations to be predicted on the leeward side. 2.5 Predicted concentration values ​​to achieve PM2.5 concentration separation in target areas 2.5 Accurate prediction of concentration.

[0139] Spatial location refers to the area within the target region (such as the South Main Road area of ​​Guangzhou), where PM2.5 needs to be calculated using the windward / leeward target attenuation prediction model. 2.5 The specific spatial location of the concentration needs to have clear spatial coordinates (such as latitude and longitude) and its attribution attribute. It must correspond to a specific LCZ block in the LCZ classification map (to obtain the morphological attenuation attribute of that location), and it must also be clear whether it belongs to the windward or leeward side (to determine the target model to be used and the corresponding distance calculation method, such as distance along the prevailing wind direction for the windward side and distance perpendicular to the road direction for the leeward side). This location is the spatial carrier of the model input data and also the PM2.5 concentration. 2.5 The final associated object of the concentration prediction value can be used for subsequent mapping of PM2.5 in the target area. 2.5 Concentration spatial distribution maps provide basic spatial units for accurately identifying high-concentration pollution sites.

[0140] For comparison of technical effectiveness, existing technologies can be used as a reference. Current research generally uses statistical regression or machine learning methods to establish PM. 2.5 The model correlates with macro-meteorological factors, but neglects the morphological attenuation effect of pollutants traversing different LCZ blocks driven by background wind fields. For example, high-density building clusters promote diffusion through enhanced turbulence, while vegetation cover may accelerate sedimentation due to adsorption, but current technologies have not yet established a physical quantitative framework for such spatially heterogeneous attenuation. Furthermore, the asymmetry of diffusion paths on the windward and leeward sides further increases the complexity of predicting concentration distribution patterns.

[0141] Therefore, an efficient prediction model integrating street morphology attenuation theory is proposed. By decoupling the attenuation mechanism of pollutant transport based on LCZ properties, this model can achieve PM2.5 attenuation under the influence of background wind field. 2.5 Accurate prediction models for the spatial distribution of concentrations urgently need to be developed.

[0142] To address the aforementioned problems, this invention provides a PM based on the theory of street morphological attenuation. 2.5 For concentration prediction methods, please refer to [link / reference]. Figure 8 S100 Data Construction: Obtain LCZ classification maps of the target area, extract the inherent attributes (building density, average height) and additional attributes of each LCZ block, and simultaneously collect meteorological data (wind direction, wind speed) and road PM2.5. 2.5 Concentration monitoring data; S200 attenuation theoretical framework construction: based on the turbulent diffusion attenuation equation, the LCZ comprehensive attenuation factor is defined, and the windward and leeward sides are divided according to the prevailing wind direction. Considering both the windward and leeward sides, PM2.5 concentrations are measured on both sides. 2.5 The influence of particulate matter concentration and distance; S300 side-by-side attenuation model establishment: based on attenuation theory, windward and leeward side models are constructed in combination with LCZ; S400 multi-level attenuation region iterative fitting: using multi-level attenuation region iterative fitting technology, the model parameters are calibrated through multi-level iteration; S500 dynamic attenuation factor output: after multi-level attenuation region iterative fitting, the pollutant attenuation factor prediction model is dynamically verified and output.

[0143] This invention establishes the relationship between the spatial morphology of urban blocks and local pollutant PM2.5 under the influence of background wind fields. 2.5 A quantitative physical correlation model for the spatial distribution pattern of PM2.5 concentrations. This model posits that PM2.5 emissions from road traffic mobile sources... 2.5 During the diffusion process, after initial emission, pollutants gradually diffuse and dilute into the surrounding environment under the influence of the background wind field. However, when traversing urban blocks with different LCZ types, the attenuation of pollutant propagation varies due to the inherent and additional properties of the LCZ, thus causing PM2.5 concentrations to rise. 2.5 The distributions differ. This is for rapid PM prediction. 2.5 This invention first selects the Nanda Expressway in Guangzhou and generates an LCZ (Localized Zone) classification map through remote sensing imagery and GIS spatial analysis; it then extracts the morphological parameters of each LCZ block, decomposing them into inherent attributes (building density, average height) and additional attributes (vegetation coverage, impermeable surface ratio); simultaneously, it acquires meteorological monitoring data (wind direction, wind speed) and PM2.5 levels around the road. 2.5 Concentration data. Then, based on inherent and additional attributes, a comprehensive LCZ attenuation factor is constructed. Simultaneously, according to the prevailing wind direction in the target area, the windward and leeward sides are divided to determine the diffusion influence coefficient. The inverse distance weighting method is used to obtain the PM2.5 concentrations on both the windward and leeward sides. 2.5The results of particulate matter concentration and distance are obtained; then, using the multi-level attenuation region iterative fitting technique, the inherent attenuation coefficient and additional attenuation coefficient of the attenuation model on the windward and leeward sides are determined; finally, the pollutant attenuation factor prediction model is dynamically verified and output.

[0144] In summary, this invention, by subdividing LCZ types and incorporating inherent attributes (such as building density and height) and additional attributes (such as vegetation cover and impermeable surface area), and specifically distinguishing the differences in pollutant diffusion on the windward and leeward sides, enables the model to accurately capture the impact mechanism of different street block morphologies on pollution diffusion, solving the problem of traditional models failing to quantify the physical mechanisms of street block attenuation. Furthermore, this invention employs multi-level iterative calculations to fit key LCZ parameters, avoiding the defects of traditional models' parameter generalization and disconnect from real-world scenarios. Through iterative optimization, parameters can dynamically match the diffusion characteristics of different LCZ types and the windward / leeward sides, significantly reducing the impact of parameter errors on prediction results. Moreover, this invention does not rely on supercomputing resources. Through the combination of refined parameter fitting and spatial analysis, it ensures real-time prediction capabilities at the city scale while enabling the prediction model to output accurate side-specific concentration data and provide clear pattern support for pollution control (such as targeted optimization of LCZ morphology in high-concentration areas), achieving a technical closed loop of "accuracy improvement - pattern revelation - practical guidance."

[0145] In this embodiment of the invention, the present invention provides a PM based on the theory of street morphological attenuation. 2.5 The concentration prediction method acquires urban morphology information and meteorological data of the target area, and outputs an LCZ classification map of the target area and the morphological attenuation attributes of each LCZ block in the LCZ classification map based on multiple preset LCZ types and urban morphology information. Based on meteorological data, multiple LCZ blocks in the LCZ classification map are classified, and multiple windward and leeward side blocks are output. The PM2.5 concentration of each windward and leeward side block is calculated. 2.5 The distance between the concentration movement test point and the road-borne mobile source was determined; based on meteorological data, the pollutant diffusion path was determined, and the windward and leeward blocks within the pollutant diffusion path were designated as target blocks. Multiple target blocks were then divided into attenuation zones, resulting in multiple windward and leeward attenuation zones. An initial attenuation prediction model for the windward and leeward sides was constructed, and the morphological attenuation attributes of the target blocks within each windward and leeward attenuation zone and PM2.5 levels were used. 2.5 The distance between the concentration movement test point and the road-based mobile source was used to perform multi-level attenuation region iterative fitting on the initial attenuation prediction model on the windward and leeward sides to determine the target attenuation prediction model on both sides. The target area was then analyzed for PM2.5 concentrations on both the windward and leeward sides using the windward and leeward side target attenuation prediction model. 2.5 Concentration prediction, outputting the PM2.5 concentration on the windward side of the target area. 2.5 Predicted concentration values ​​and leeward PM 2.5Concentration prediction; Based on the above scheme, this invention deeply integrates the LCZ classification system with the side-by-side zoning analysis logic, and accurately quantifies the impact of street blocks on PM through morphological attenuation attributes. 2.5 The diffusion attenuation mechanism fills the technical gaps in traditional models that do not associate LCZ attributes and do not quantify the attenuation effect of blocks. At the same time, multi-level iterative fitting does not rely on supercomputing resources and complex parameter inputs, which simplifies the model's computational cost and allows the model to fully characterize the pollution diffusion patterns under different spatial dimensions through refined fitting of sides (windward / leeward) and levels (attenuation areas), significantly improving prediction accuracy.

[0146] Reference Figure 9 The target attenuation prediction model on the windward and leeward sides is applied to output the PM on the windward side. 2.5 Predicted concentration values ​​and leeward PM 2.5 The present invention further includes the following steps: (Analysis of the predicted concentration values ​​is performed.)

[0147] Step 901: Using the ordinary Kriging method, based on the PM levels on the windward side... 2.5 Predicted concentration values ​​and leeward PM 2.5 Spatial interpolation of concentration prediction values ​​is performed to output PM2.5. 2.5 Concentration field.

[0148] Step 902: Spatial autocorrelation analysis and hotspot analysis are used to analyze PM. 2.5 The concentration field is analyzed, and the spatial autocorrelation analysis results and hotspot analysis results are output.

[0149] Step 903: Use kernel density estimation method to analyze PM on the windward side. 2.5 Predicted concentration values ​​and leeward PM 2.5 The predicted concentration values ​​are analyzed, and the results of the nuclear density analysis on the windward and leeward sides are output.

[0150] The spatial autocorrelation analysis results include the global Moran index and the Anselin local Moran index.

[0151] It should be noted that PM is generated using Kriging interpolation. 2.5 Concentration field, where, in the Kriging interpolation step, the ordinary Kriging formula is specifically: ;

[0152] ;

[0153] This formula is used to generate spatially continuous PM. 2.5 Concentration distribution field (PM) 2.5 Concentration field). Among them, Spatial location PM at the location 2.5 Concentration interpolation predictions; multiple such interpolation values ​​can form a spatially connected PM2.5 concentration prediction.2.5 Concentration field; These are the interpolation weighting coefficients at the i-th spatial location, used to measure the relationship between each sample point and its spatial location. The contribution of the concentration field at a given location, and its weight, are typically proportional to the sum of the sample points. The spatial distance and spatial correlation between them are related factors; In spatial location PM at the location 2.5 Concentration attribute value, this value specifically refers to , ; To represent PM participating in the interpolation calculation 2.5 The number of concentration attribute values, i.e. the total number of spatial location sample points used to generate the concentration field.

[0154] The interpolation weighting coefficients are as follows:

[0155] ;

[0156] in, Indicates a known observation point (spatial location) and spatial location The covariance function values ​​between ) and ) Known observation points Interpolation point The covariance function values ​​between them. Based on the prediction results of this model (i.e.) , The spatial distribution was calculated.

[0157] Furthermore, in the spatial autocorrelation analysis step, this invention introduces the global Moran index and the Anselin local Moran index to evaluate the spatial clustering pattern of the predicted concentration field. The formula for the global Moran index is:

[0158] ;

[0159] in, The global Moran index is used to measure PM2.5 levels within the study area. 2.5 The overall spatial autocorrelation of concentration (or other spatial properties) generally ranges from [value range missing]. Between these values, a positive value indicates positive spatial autocorrelation (clustered distribution of similar attribute values), a negative value indicates negative spatial autocorrelation (adjacent distribution of dissimilar attribute values), and a value close to 0 indicates spatial randomness. The elements in the spatial weight matrix represent the spatial weight relationship between spatial position i and spatial position j, which are used to quantify the spatial proximity between two points. PM for all sample points (spatial locations) 2.5The average value of the concentration attribute; , PM at spatial locations i and j respectively 2.5 Concentration attribute values, specifically referring to the values ​​predicted by this model using the aforementioned Kriging interpolation step (i.e., concentration attribute values). , ) Output continuous concentration field The values ​​taken on the corresponding spatial units are derived from the concentration field generated after interpolation. n represents the total number of sample points participating in the spatial autocorrelation analysis within the study area;

[0160] The formula for the Anserin local Moran index is:

[0161] ;

[0162] in, The local Moran index at spatial location i is used to measure the degree of local spatial autocorrelation between this location and its surrounding area. It can identify spatial patterns of local clustering (high-high, low-low) or local anomalies (high-low, low-high).

[0163] Furthermore, in the hotspot analysis step, this invention introduces the Getis-Ord Gi* index, the formula of which is:

[0164] ;

[0165] in, Let Gi* be the Getis-Ord index at spatial location i, representing the thermal analysis result; The total number of regional features, region It is the spatial weight between regional features i and j. It is the attribute value of the region, and in this invention, it specifically refers to the PM2.5 concentration value at location j, that is, the concentration field grid value obtained after spatial interpolation.

[0166] Furthermore, in the kernel density estimation step, this invention introduces kernel density for analyzing PM. 2.5 The formula for the spatial distribution density of concentration is:

[0167] ;

[0168] in, Spatial location PM at the location 2.5 The concentration kernel density estimate represents the results of kernel density analysis, including both windward and leeward kernel density analyses. This result reflects the PM2.5 concentration in the surrounding area at that location. 2.5Spatial distribution density of concentration; the higher the density value, the stronger the concentration aggregation in that area. Bandwidth (smoothing parameter) controls the smoothness of the kernel function and determines the range of influence of each data point on the surrounding space. The larger the bandwidth, the smoother the density distribution; conversely, the more detailed the data. For the i-th PM 2.5 The spatial coordinates of the concentration sample point, and the corresponding concentration value. , ; The kernel function is used to estimate the position. With the location of the sample point Distance-based weighting; concentration corresponding to the sample point , It has been incorporated into the previous data preprocessing (i.e., the spatial location of each sample point). (Latitude and longitude) have already been compared with the PM at that location in the previous period. 2.5 The concentration predictions correspond one-to-one, and the kernel density estimation reflects the spatial clustering of concentrations through location weights.

[0169] It is worth mentioning that spatial autocorrelation analysis results (including global and local Moran indices) can help managers grasp the overall distribution patterns and local clustering characteristics of pollution—if the global Moran index shows a significant positive correlation, it indicates that PM2.5 levels in the target area are relatively high. 2.5 Pollution exhibits a concentrated distribution rather than random diffusion, allowing for the priority development of regional integrated control plans. "High-high" clusters (adjacent high-concentration areas) identified by the Moran's index are the core areas for pollution control, requiring focused resource allocation (e.g., strengthening industrial emission reduction and strictly controlling mobile source emissions). "Low-low" clusters can serve as benchmarks for pollution control effectiveness, allowing for analysis of their low-pollution causes (e.g., high vegetation cover, appropriate LCZ morphology) and the dissemination of best practices. Furthermore, comparing local autocorrelation results at different times can assess the effectiveness of phased control measures (e.g., a reduction in the size of "high-high" clusters indicates effective measures). Hotspot analysis results (e.g., the Getis-OrdGi* index) can accurately pinpoint PM levels. 2.5The application of high-concentration "hotspot areas" (statistically significant high-value clusters) and low-concentration "coldspot areas" focuses more on refined management. For hotspot areas, fixed monitoring stations and mobile sensors can be deployed more densely to track pollution dynamics in real time and initiate emergency responses (such as temporarily controlling traffic flow in the surrounding area during pollution peaks). At the same time, potential pollution contributors within hotspot areas (such as restaurant fumes and construction dust) can be investigated. Coldspot areas can serve as a reference for "ecological buffer zone" planning. By preserving or adding green spaces and permeable surfaces, the interception and sedimentation of pollution from surrounding hotspot areas can be strengthened. In addition, spatial correlation analysis between hotspot areas and coldspot areas (such as whether the downwind direction of a hotspot area is a coldspot area) can also help determine the impact of prevailing winds on pollution diffusion, providing a basis for the layout adjustment of wind-sensitive industries (such as air-polluting enterprises). The kernel density analysis results on the windward and leeward sides, by distinguishing the differentiated impact of wind direction on pollution, can provide in-depth support for urban morphology optimization and regional governance. By comparing the distribution of high-value kernel density areas on both sides, if the high-value kernel density areas on the windward side extend in a strip along the prevailing wind direction, it indicates that the prevailing wind drives pollution diffusion. Low-lying vegetation belts or ventilation corridors should be planned in the upstream area of ​​the windward side (such as along roadsides) to reduce pollution transmission downstream. If the high-value kernel density areas on the leeward side are concentrated in densely built-up LCZ blocks (such as compact high-rise buildings), it indicates that the leeward side is more susceptible to pollution. Building obstruction creates eddies, leading to pollution retention. Optimizing the LCZ morphology in this area (e.g., reducing building density and increasing open space) is necessary to improve ventilation. Simultaneously, the lateral core density results can guide the differentiated layout of the monitoring network. For example, adding monitoring points in areas with large core density gradients on the windward side (rapid concentration decay sections) can capture pollution diffusion patterns. Conversely, focusing on monitoring the interaction between buildings and pollution in high-density areas on the leeward side provides measured data for correcting the LCZ morphology decay properties, ultimately achieving "site-specific and wind-differentiated" PM2.5 monitoring. 2.5 Precise pollution control and optimized urban planning.

[0170] In this embodiment of the invention, the present invention subdivides LCZ types (considering inherent / additional attributes) and their impact on pollutant diffusion (especially on the windward / leeward sides), employs multi-level iterative calculations to fit key LCZ parameters, and constructs an attenuation model to correct the influence of street block morphology. Combining ArcGIS spatial analysis (Kriging interpolation, hotspot analysis, spatial autocorrelation) and iterative calibration techniques, the PM2.5 concentrations under varying wind direction and distance are revealed. 2.5 The concentration patterns were studied, and a prediction model was established.

[0171] Please see Figure 10 , Figure 10 This invention provides a PM based on the street morphology attenuation theory, as shown in Embodiment 3. 2.5 Block diagram of the concentration prediction device.

[0172] This invention provides a PM based on the theory of street morphological attenuation.2.5 Concentration prediction device, including:

[0173] The acquisition module 1001 is used to acquire urban morphology information and meteorological data of the target area, and output the LCZ classification map of the target area and the morphological attenuation attributes of each LCZ block in the LCZ classification map according to multiple preset LCZ types and urban morphology information.

[0174] Output module 1002 is used to classify multiple LCZ blocks in the LCZ classification map based on meteorological data and output multiple windward and leeward blocks.

[0175] Calculation module 1003 is used to calculate the PM2.5 concentration in each windward and leeward block based on the meteorological data. 2.5 The distance between the concentration movement test point and the road-based mobile source;

[0176] The segmentation module 1004 is used to determine the pollutant diffusion path based on meteorological data, and to take the windward and leeward blocks within the pollutant diffusion path as target blocks, and to divide the multiple target blocks into attenuation areas to obtain multiple windward and leeward level attenuation areas.

[0177] Iteration module 1005 is used to construct the initial attenuation prediction model for the windward and leeward sides, and adopts the morphological attenuation attributes and PM of the target blocks in the attenuation regions of each windward and leeward side level. 2.5 The distance between the concentration movement test point and the road movement source is used to perform multi-level attenuation region iterative fitting on the initial attenuation prediction model on the windward and leeward sides to determine the target attenuation prediction model on the windward and leeward sides.

[0178] Prediction module 1006 is used to predict PM levels on the windward and leeward sides of a target area using a target attenuation prediction model on both sides. 2.5 Concentration prediction, outputting the PM2.5 concentration on the windward side of the target area. 2.5 Predicted concentration values ​​and leeward PM 2.5 Predicted concentration values.

[0179] In one optional device embodiment, it further includes:

[0180] The first module is used to apply the ordinary Kriging method to the PM on the windward side. 2.5 Predicted concentration values ​​and leeward PM 2.5 Spatial interpolation of concentration prediction values ​​is performed to output PM2.5. 2.5 Concentration field;

[0181] The second module is used to analyze PM using spatial autocorrelation analysis and hotspot analysis methods, respectively. 2.5 The concentration field is analyzed, and the spatial autocorrelation analysis results and hotspot analysis results are output.

[0182] The third module is used to perform kernel density estimation on the windward side PM. 2.5 Predicted concentration values ​​and leeward PM 2.5 The predicted concentration values ​​are analyzed, and the results of the nuclear density analysis on the windward and leeward sides are output.

[0183] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described device and module can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0184] This invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program; when the computer program is executed by the processor, the processor performs PM based on the street morphology attenuation theory as described in any of the above embodiments. 2.5 Steps in concentration prediction methods.

[0185] This invention also provides a computer-readable storage medium storing a computer program / instructions thereon, which, when executed by a processor, implements PM based on the street morphology attenuation theory as described in any of the above embodiments. 2.5 Steps in concentration prediction methods.

[0186] This invention also provides a computer program product, including a computer program / instructions, which, when executed by a processor, implements a PM based on the street morphology attenuation theory as described in any of the above embodiments. 2.5 Steps in concentration prediction methods.

[0187] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A PM based on the theory of street morphology attenuation 2.5 Concentration prediction method, characterized in that, include: Acquire urban morphology information and meteorological data of the target area, and output an LCZ classification map of the target area and the morphological attenuation attributes of each LCZ block in the LCZ classification map based on multiple preset LCZ types and the urban morphology information. Based on the meteorological data, multiple LCZ blocks in the LCZ classification map are classified, and multiple windward and leeward blocks are output. Based on the meteorological data, the PM2.5 concentration in each of the windward and leeward blocks is calculated. 2.5 The distance between the concentration movement test point and the road-based mobile source; Based on the meteorological data, the pollutant diffusion path is determined, and the windward and leeward blocks within the pollutant diffusion path are taken as target blocks. The multiple target blocks are then divided into attenuation areas to obtain multiple windward and leeward level attenuation areas. An initial attenuation prediction model is constructed for both the windward and leeward sides, and the morphological attenuation attributes and PM of target blocks in each of the windward and leeward attenuation level regions are used. 2.5 The distance between the concentration movement test point and the road movement source is used to perform multi-level attenuation region iterative fitting on the initial attenuation prediction model on the windward and leeward sides to determine the target attenuation prediction model on the windward and leeward sides. The target area is analyzed for PM levels on the windward and leeward sides using the aforementioned target attenuation prediction model. 2.5 Concentration prediction, outputting the PM2.5 concentration on the windward side of the target area. 2.5 Predicted concentration values ​​and leeward PM 2.5 Predicted concentration values.

2. PM based on street morphological attenuation theory as described in claim 1 2.5 Concentration prediction method, characterized in that, The urban morphology information includes road network data, remote sensing images, and surface parameter data; the step of outputting an LCZ classification map of the target area and the morphological attenuation attributes of each LCZ block in the LCZ classification map based on multiple preset LCZ types and the urban morphology information includes: The road network data is preprocessed to output the urban road network of the target area; The urban road network of the target area is used as the LCZ block segmentation boundary. Based on the LCZ block segmentation boundary, the target area is segmented according to the remote sensing data using GIS spatial analysis methods, and multiple initial LCZ blocks are output. Extract the morphological attenuation attributes of each initial LCZ block, and classify each initial LCZ block according to the morphological attenuation attributes of each initial LCZ block based on multiple preset LCZ types, and output multiple LCZ blocks with preset LCZ type labels. Each LCZ block with a preset LCZ type label is verified to generate multiple LCZ blocks, and multiple LCZ blocks are used to construct an LCZ classification map of the target area.

3. PM based on street morphology attenuation theory as described in claim 1 2.5 Concentration prediction method, characterized in that, The windward and leeward side blocks include windward and leeward side blocks; the classification of multiple LCZ blocks in the LCZ classification map based on the meteorological data outputs multiple windward and leeward side blocks, including: The meteorological data is used to divide the target area into wind-side partitions, and the windward and leeward global partition ranges of the target area are output. Based on the global partition range on the windward side and the global partition range on the leeward side, the area corresponding to each LCZ block is used to classify each LCZ block on the windward side, thereby determining multiple windward side blocks and multiple leeward side blocks.

4. PM based on street morphological attenuation theory as described in claim 3 2.5 Concentration prediction method, characterized in that, The PM 2.5 The distance between the concentration movement test point and the road-based mobile source includes PM2.5 concentration. 2.5 Distance between concentration testing point and road mobile source along the prevailing wind direction, PM 2.5 The vertical road distance between the concentration testing point and the road-borne mobile source; based on the meteorological data, the PM2.5 concentration in each windward and leeward block is calculated. 2.5 The distance between the concentration movement test point and the road-based mobile source includes: PM2.5 samples were collected from the windward and leeward sides of each of the aforementioned blocks. 2.5 Geographic latitude and longitude coordinates of concentration movement test points and representative geographic latitude and longitude coordinates of road movement sources; The spherical distance calculation formula is used based on the PM. 2.5 Using the geographic latitude and longitude coordinates of the concentration movement test point and the representative geographic latitude and longitude coordinates of the road movement source, PM2.5 is calculated. 2.5 The baseline Euclidean distance between the geographic latitude and longitude coordinates of the concentration movement test point and the representative geographic latitude and longitude coordinates of the road movement source; Based on the meteorological data, the standard azimuth of the prevailing wind direction is determined, and the PM is calculated. 2.5 The azimuth angle of the line connecting the concentration movement test point relative to the road movement source; Calculate the angle between the azimuth of the connecting line and the standard azimuth of the prevailing wind direction, and multiply the reference Euclidean distance by the cosine of the angle between the connecting line and the standard azimuth of the prevailing wind direction to obtain the PM. 2.5 Distance between the concentration testing point and the road-borne mobile source along the prevailing wind direction; Calculate the PM 2.5 The angle between the line connecting the concentration movement test point and the road movement source and the direction perpendicular to the road, and the reference Euclidean distance is used to calculate the PM2.5 concentration. 2.5 The PM2.5 concentration is obtained by multiplying the cosine of the angle between the line connecting the concentration movement test point and the road movement source and the direction perpendicular to the road. 2.5 The vertical road distance between the concentration test point and the road-moving source.

5. PM based on street morphological attenuation theory as described in claim 4 2.5 Concentration prediction method, characterized in that, The windward and leeward level attenuation regions include windward level attenuation regions and leeward level attenuation regions; the windward and leeward initial attenuation prediction models include windward initial attenuation prediction models and leeward initial attenuation prediction models; the morphological attenuation attributes and PM of target blocks in each of the windward and leeward level attenuation regions are used. 2.5 The distance between the concentration movement test point and the road movement source is used to perform multi-level attenuation region iterative fitting on the initial attenuation prediction model on the windward and leeward sides to determine the target attenuation prediction model on the windward and leeward sides, including: Based on the morphological attenuation properties of the target blocks in the N-level attenuation region within each of the aforementioned windward attenuation levels and PM 2.5 The distance between the concentration test point and the road mobile source along the prevailing wind direction is used to fit the Nth group of windward LCZ parameters, where N is the number of the windward attenuation level region, and N≥1; The windward LCZ parameters of the Nth group are set as known quantities, and combined with the morphological attenuation properties of the target blocks in the N+1 level attenuation region of each windward attenuation region and PM 2.5 The distance between the concentration test point and the road-borne mobile source along the prevailing wind direction is used to fit the N+1th group of windward LCZ parameters. If the currently fitted windward level attenuation region is the last windward level attenuation region that was not fitted, and the windward LCZ parameter corresponding to the windward level attenuation region is fitted, the windward LCZ parameter obtained from this fitting is used as the windward attenuation factor matrix. The initial attenuation prediction model on the windward side is updated using the windward attenuation factor matrix, and the target attenuation prediction model on the windward side is output. Based on the morphological attenuation properties of the target blocks in the M-level attenuation region within each of the aforementioned leeward attenuation levels and PM 2.5 The vertical road distance between the concentration test point and the road mobile source is used to fit the Mth group of leeward LCZ parameters, where M is the sequence number of the leeward attenuation level region, and M≥1; The leeward LCZ parameters of the Mth group are set as known quantities, and combined with the morphological attenuation properties of the target blocks in the M+1 level attenuation region of each leeward level attenuation region and PM 2.5 The vertical road distance between the concentration test point and the road-borne mobile source is used to fit the leeward LCZ parameters of the (M+1)th group. If the currently fitted leeward-side level attenuation region is the last leeward-side level attenuation region that was not fitted, and after the leeward-side LCZ parameters corresponding to the leeward-side level attenuation region are fitted, the leeward-side LCZ parameters obtained from this fitting are used as the leeward-side attenuation factor matrix. The leeward attenuation factor matrix is ​​used to update the initial leeward attenuation prediction model, and the target leeward attenuation prediction model is output. Specifically, the initial attenuation prediction model on the windward side is as follows: ； In the formula, PM on the windward side 2.5 Concentration prediction; The attenuation factor for the target block with LCZ type i is preset in the windward attenuation region; , , , , , The LCZ parameters on the windward side are to be fitted to the initial attenuation prediction model on the windward side. The number of preset LCZ types that the pollutants pass through on the windward diffusion path in the diffusion path; Set the sky angle coefficient factor for the target block with LCZ type i in the windward attenuation region; The aspect ratio of the street canyon of the target block with LCZ type i preset in the windward attenuation region; The roughness element height of the target block with LCZ type i is preset in the windward attenuation region; Preset the building density of the target block with LCZ type i in the windward attenuation region; The impermeable surface ratio of the target block with LCZ type i is preset in the windward attenuation region; The permeable surface ratio of the target block with LCZ type i is preset in the windward attenuation region; For PM 2.5 Distance between the concentration testing point and the road-borne mobile source along the prevailing wind direction; Preset the emission intensity of traffic mobile sources with LCZ type i in the windward attenuation zone; The initial attenuation prediction model on the leeward side is specifically as follows: ； In the formula, leeward side PM 2.5 Concentration prediction; The number of preset LCZ types that the pollutant passes through on the leeward side of the diffusion path; The attenuation factor for the target block with LCZ type i is preset in the leeward attenuation region; , , , , , The leeward LCZ parameters to be fitted for the initial attenuation prediction model on the leeward side; Set the sky angle coefficient factor for the target block with LCZ type i in the leeward attenuation region; The aspect ratio of the street canyon of the target block with LCZ type i preset in the leeward level attenuation region; The roughness element height of the target block with LCZ type i is preset in the leeward level attenuation region; Preset the building density of the target block with LCZ type i in the leeward attenuation region; The impermeable surface ratio of the target block with LCZ type i is preset in the leeward level attenuation region; The permeable surface ratio of the target block with LCZ type i is preset in the leeward-side attenuation region; For PM 2.5 The vertical road distance between the concentration testing point and the road-borne mobile source; The emission intensity of traffic mobile sources with LCZ type i is preset in the leeward-side level attenuation region.

6. PM based on street morphological attenuation theory as described in claim 1 2.5 Concentration prediction method, characterized in that, Also includes: Using the ordinary Kriging method based on the windward side PM 2.5 Predicted concentration values ​​and the leeward side PM 2.5 Spatial interpolation of concentration prediction values ​​is performed to output PM2.

5. 2.5 Concentration field; Spatial autocorrelation analysis and hotspot analysis were used to analyze the PM. 2.5 The concentration field is analyzed, and the spatial autocorrelation analysis results and hotspot analysis results are output. The kernel density estimation method was used to analyze the PM on the windward side. 2.5 Predicted concentration values ​​and the leeward side PM 2.5 The predicted concentration values ​​are analyzed, and the results of the nuclear density analysis on the windward and leeward sides are output.

7. A PM based on the theory of street morphology attenuation 2.5 Concentration prediction device, characterized in that, include: The acquisition module is used to acquire urban morphology information and meteorological data of the target area, and output the LCZ classification map of the target area and the morphological attenuation attributes of each LCZ block in the LCZ classification map according to multiple preset LCZ types and the urban morphology information. The output module is used to classify multiple LCZ blocks in the LCZ classification map based on the meteorological data and output multiple windward and leeward blocks. The calculation module is used to calculate the PM2.5 concentration in each of the windward and leeward blocks based on the meteorological data. 2.5 The distance between the concentration movement test point and the road-based mobile source; The segmentation module is used to determine the pollutant diffusion path based on the meteorological data, and to take the windward and leeward blocks within the pollutant diffusion path as target blocks, and to divide the multiple target blocks into attenuation regions to obtain multiple windward and leeward level attenuation regions. The iterative module is used to construct the initial attenuation prediction model on the windward and leeward sides, and adopts the morphological attenuation attributes and PM of the target blocks in each of the windward and leeward attenuation level regions. 2.5 The distance between the concentration movement test point and the road movement source is used to perform multi-level attenuation region iterative fitting on the initial attenuation prediction model on the windward and leeward sides to determine the target attenuation prediction model on the windward and leeward sides. The prediction module is used to predict PM levels on the windward and leeward sides of the target area using the windward and leeward side target attenuation prediction model. 2.5 Concentration prediction, outputting the PM2.5 concentration on the windward side of the target area. 2.5 Predicted concentration values ​​and leeward PM 2.5 Predicted concentration values.

8. A computer device, characterized in that, The system includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, it causes the processor to perform PM based on the street morphology attenuation theory as described in any one of claims 1-6. 2.5 Steps in concentration prediction methods.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed, it implements the PM based on the street morphology attenuation theory as described in any one of claims 1-6. 2.5 Concentration prediction methods.

10. A computer program product, characterized in that, The computer program product includes a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions, wherein, when the program instructions are executed by a computer, the computer performs the PM based on the street morphology attenuation theory as described in any one of claims 1-6. 2.5 Concentration prediction methods.

Images